Incorporating Public Values Through Multiple Accountability: A Case Study on Quality Regulation of Emergency Care in the Netherlands by an Independent Regulatory Agency

In this paper, we explore how multiple accountability (MA) can enable an independent regulatory agency to deal with multiple conflicting public values in a complex and politically salient decision-making process. We examined the decision-making process of the Dutch National Health Care Institute on quality regulation of emergency care in the Netherlands. Using insights derived from ethnography, document analysis, and interviews, we show that MA resulted from strategic interactions between the Institute’s vertical and horizontal accountability forums. We argue that MA impeded efficiency but also enabled the Institute to deal with multiple conflicting public values.


Introduction
Since the 1980s, the privatization and reregulation of public services in the context of new public management reforms has led to a shift from government to systems of dispersed governance in many European countries (Bovens & Schillemans, 2014;Majone, 1994).These developments have restricted hierarchical influence and have sparked a tendency to introduce new accountability practices to overcome accountability deficits in policymaking (Schillemans, 2011).A large body of accountability research focuses on the public accountability of independent regulatory agencies (IRAs), to which governments have increasingly delegated tasks (Helderman et al., 2012;Majone, 1994).As ministerial control on these agencies is limited, they are not affected by traditional democratic accountability through elections (Durose et al., 2015).
This accountability deficit is particularly perceived as pressing since agencies often execute substantive tasks.Therefore, they give a certain meaning to political values when making decisions and these analyses cannot be laid down in legal procedures (Bach & Jann, 2010;Eriksen, 2021;Majone, 1996).Although much research has been done on value conflict, finding better ways to deal with different values remains a pressing issue for policymakers (Kernaghan, 2003).Besides coping with different (often conflicting) values, agencies must deal with numerous accountability practices because of their accountability deficit.This phenomenon is called multiple accountability (MA).Scholars have largely focused on the drawbacks of MA, such as high costs, pressure on public officials, politicization, and confusion (Flinders, 2011;Koppell, 2005;Willems & Van Dooren, 2012).
However, the benefits of MA, such as increased reliability of oversight and reduction of information asymmetry, have also been investigated (Schillemans, 2010).In our ethnographic study, we will focus on the benefit addressed by both Scott (2000) and Schillemans (2010), who state that MA balances different values because different forums have competing agendas, concerns, powers, procedures, and capacities.Previous empirical research that addresses how organizations and individuals deal with competing values and accountability shows the often complex, messy, and political nature of these processes.Oldenhof et al. (2014) conducted observations and interviews to show how public managers use justifications to deal with value conflicts in their daily work.Brunsson (1989, pp. 4-9) finds that, to survive, organizations strategically try to gain legitimacy through the creation of structures, processes, and ideologies which reflect the inconsistent norms in their environment.The well-known garbage can process that Cohen et al. (1972) describe is more coincidental.It stresses that outcomes are determined by complex interactions between streams of choices, problems, solutions, and participants which come together in a metaphorical garbage can.
A large body of publications from PA scholars from different countries acknowledges the relevance of the public accountability deficit and the role of agencies in safeguarding public values (Overman et al., 2015).Despite this, specific research on how agencies deal with both public accountability and public values is limited.Schillemans (2011) is a notable exception who studied accountability forums of nine Dutch agencies.He shows that MA benefited these agencies in several ways, including the incorporation of different legitimate values into their decision-making process (Schillemans, 2010).Our case study builds on his work by closely examining the day-to-day decisions in a single policymaking process of a single agency.Although ethnography in PA research dates to the work of Heclo and Wildavsky (1974) and its value is stressed (Herzog & Zacka, 2019;Rhodes, 2014), in depth qualitative analyses using ethnography on agency accountability are uncommon.Our ethnographic approach allows us to show the complex accountability and value dynamics in detail to further explore the possible benefit(s) of MA for semi-independent agencies.
Our aim is to explore how MA can help an agency to deal with multiple public values during a complex and salient policymaking process.Our case study was the development of the Dutch national quality standard for emergency care.We studied the role of the National Health Care Institute (Zorginstituut Nederland, ZIN; hereafter referred to as "the Institute"), an influential IRA in the Netherlands.From May 2019 until September 2019, we closely followed the process through participant observations, interviews, and document analysis.A related societal and political discussion about the trade-off between the public values quality, affordability and accessibility of nationwide emergency care in the Netherlands made this a salient process.As in many other countries, policymakers in the Netherlands have focused on nationwide concentration of emergency care to reduce costs and improve quality in the past two decades.However, whether this nationwide concentration has reduced costs and improved quality has been disputed.In addition, the decline in number of emergency departments (EDs) from 105 in 2010 to 87 in 2016 due to mergers has raised questions about regional accessibility of emergency care (Gaakeer et al., 2018).In this paper, we take the three public values of care-quality, affordability, and accessibility-as a starting point because these are the three formal pillars of the Dutch healthcare system which the Institute is expected to promote (Helderman et al., 2014, p. 91).In our analysis, we show how these values were continuously given meaning and weight by the actors involved.
First, we conceptualize multiple public accountability and public values.Second, we explain how we selected our case and collected and analyzed our data.Third, we discuss our empirical findings on the development of the national quality standard for emergency care in the Netherlands.Fourth, we draw our conclusions and discuss how MA allowed the Institute to consider and appraise multiple conflicting public values.

Multiple Public Accountability and Multiple Public Values
Dealing with multiple public accountability (Schillemans, 2010) and conflicting public values (Kernaghan, 2003) are common issues for policymakers today.On the one hand, agencies increasingly deal with different public values, which raises accountability problems (Eriksen, 2021).On the other hand, multiple public values are produced through MA practices (Schillemans, 2010;Scott, 2000).In this paper, we want to understand accountability practices and the role of public values in these practices.

Multiple Public Accountability
The concept accountability can be broadly divided into two categories First, since accountability is seen as a central element of democracy, it has often been defined as a virtue of good governance (Bovens, 2010).Second, accountability is commonly understood as a communicative interaction.Bovens (2007) defined it as "a relationship between an actor and a forum, in which the actor has an obligation to explain and justify his or her conduct, the forum can pose questions and pass judgement, and the actor may face consequences" (Bovens, 2007, p. 450).In this paper, we use the definition by Bovens (2007) since our aim is to analyze how MA is constituted in practice.His definition allows for taking a broader lens on the concept beyond hierarchical principal-agent ideas of account holding and formal sanctions (Brummel, 2021;Willems & Van Dooren, 2012).Our case also shows the manifold types, forums and functions of accountability such as voluntary (Koop, 2014), felt (Overman et al., 2020), learning-oriented, and reputationbased accountability (Busuioc & Lodge, 2015;Maggetti & Papadopoulos, 2018).We place particular focus on public accountability.Here, "public" refers to the openness or accessibility of the agency's accountability to citizens and to the public sector in which the agency operates (Bovens, 2005(Bovens, , 2007)).We treat accountability as an empirical rather than as a normative concept in the sense that we do not aim to argue that there should be more or fewer accountability practices.
Vertical accountability or the central control of hierarchically superior actors such as parent departments and parliaments is usually seen as the traditional form (Schillemans, 2010).Examples of vertical accountability practices are performance indicators (Pollitt, 2006), HRM control (van Thiel & Yesilkagit, 2014), audits, annual reports (Thatcher, 2002), and ministerial questions about the agency's conduct (van Thiel & Yesilkagit, 2011).While vertical accountability mechanisms remained influential or expanded over time (Bovens, 2010;Schillemans & van Twist, 2016), forms of horizontal accountability toward parallel forums have been increasingly introduced (Schillemans, 2010).These are non-hierarchical accountability forums such as independent evaluators, the media, professional peers (Thatcher, 2002), boards of stakeholders or commissioners, clients, and interest groups (Jacobs & Schillemans, 2016).Because of this sedimentation process, agencies need to operate in seemingly redundant accountability networks of overlapping accountability practices that focus on the same topics, produce similar information, or steer agency behavior in similar directions (Scott, 2000).
The legally established vertical and horizontal accountability practices can be complemented with strategically initiated ones.First, the interest of horizontal forums in the agency's conduct can motivate strategic actions such as trying to influence policy decisions through the media or the ministry.Since horizontal forums often cannot use formal sanctions, activating an agency's vertical forum is a powerful tool for steering policy decisions (Schillemans, 2008).Second, besides being held accountable, agencies themselves render account for strategic reasons such as building trust, credibility or reputation, and gaining autonomy or resource benefits (Brummel, 2021;Schillemans & Busuioc, 2015).

Benefits and Drawbacks of Multiple Accountability
The introduction of new accountability forums and practices is commonly seen as a desirable development (Flinders, 2011).Multiple and redundant accountability can be beneficial in common dispersed governance settings where complexity and uncertainty are influential, which also applies to our case (Braithwaite, 1999;Schillemans, 2010;Scott, 2000;Willems & Van Dooren, 2012).According to Scott (2000), the benefit of redundancy is that if one accountability practice fails, the other one can still prevent the risk of unwanted behavior.Also, competing interests of accountability forums decrease the information asymmetry between agencies and their forums.The MA mechanisms all produce information which can be used as input for accountability processes of other forums (Scott, 2000).
However, scholars have also pointed out the downsides of MA, such as being costly and time-consuming and causing confusion (Schillemans, 2010).This relates to the accountability dilemma, which acknowledges that accountability practices might hamper factors necessary for effective service delivery such as freedom to manage, long-term planning, innovation, flexibility, and risk-taking (Cohen et al., 1972;Flinders, 2011, p. 600).The capacity to credibly commit to dealing with complex policy issues justifies the delegation of tasks to IRAs (Majone, 2001), which raises the question of whether MA is desirable from a public perspective.Another possible drawback is the politicization of accountability which means that accountability practices and information are used to fit partial interests.Because of this, accountability practices often focus on portraying politicians and policymakers as untrustworthy rather than on other purposes which negatively affects public trust (Flinders, 2011).The increased complexity of MA has also been addressed as worrying.Koppell (2005, p. 3) argues that organizations that try to meet conflicting expectations of accountability forums can become dysfunctional.They risk "pleasing no one while trying to please everyone," which he referred to as a multiple accountabilities disorder (MAD).

Multiple Public Values
The advantage of MA that we focus on here is that it might allow an agency to consider different and legitimate values when making decisions.According to (Scott, 2000), within the redundancy model, different mechanisms check different values.Because forums have different concerns, powers, procedures, and capacities and because they often have competing agendas, these values are balanced (Scott, 2000).Schillemans (2010) argued that different forums can safeguard different values particularly well when relevant stakeholders have diverging interests and opinions.In our study, we explore this possible benefit of MA in detail.
Public values can be defined in different ways.According to De Bruin and Dicke (2006), a value is public not private if a collective or an aggregation level can benefit from the protection of that value.However, Eriksen (2021, p. 3) referred to these values as "political values."He argues that agencies might unreflectively apply a value such as social justice to a partisan or sectarian paradigm such as liberal economic ideology.He argues that instead, agencies need to acknowledge that such values are what Gallie (1956) coined as "essentially contested concepts."Although these values are collective, they can be interpreted differently and are not viewed as positive by all members of society (Bozeman, 2007, p. 7).Although agencies always interpret values when executing political tasks, they frequently try to remain neutral by not taking a political or partisan stance and their decisions are often mistakenly not seen as political (Eriksen, 2021).
In this study, we do not focus on a singular public value (cf Moore, 2014), but rather on plural public values (de Graaf et al., 2016;Jørgensen & Bozeman, 2007).We use the definition of Thatcher and Rein (2004, p. 460) who define values as "the ultimate ends of public policy-the goals and obligations that policy aims to promote as desirable in their own right, not just as means to some other objective."We follow their idea that public values are often incommensurable, that actors cannot translate different goals in a common, overarching metric of often financial value and that there are other rational approaches to dealing with value conflict than trade-off strategies.We take a sociological perspective and analyze the decision-making of the Institute as a continuous process to show how different values were continuously made visible.This approach differs from long-term institutional approaches like cycling (sequentially emphasizing values) and firewalls (distributing values among different institutions; Thatcher & Rein, 2004).Finally, we see accountability as a public value in itself.Like de Graaf et al. ( 2016) and de Graaf and van der Wal (2010), we will show how this value can conflict with effectiveness and efficiency.

Case Description and Selection
The Institute is an interesting IRA for studying the relation between public values and MA since it is legally expected to make decisions about conflicting public values when national stakeholders are unable to compromise.It operates in the context of the Dutch healthcare system.This system is Bismarckian, which means that it is a system of health insurance rather than a Beveridge national health system funded through general taxation such as the NHS in the United Kingdom (Guy, 2016, p. 6).In 2006, the Dutch system was reformed to a system of regulated competition through the introduction of market arrangements.Since the reform, there is competition between health insurance companies and between healthcare providers.The system is based on equal access and solidarity and the market is largely regulated (Den Exter, 2010;Maarse et al., 2016).The Dutch healthcare system provides an interesting case to study dynamics between vertical and horizontal accountability forums.Unlike countries with hierarchical and relatively centralized governments such as in the United Kingdom, The Netherlands has a tradition of consensus-based democracy known as "the poldermodel."Therefore, societal actors historically have an influential authorized position as horizontal accountability forums of state actors, like the Institute (Bekker et al., 2018;Helderman et al., 2014).
The National Health Care Institute is an influential IRA in the Dutch healthcare system with a broad mandate.It supervises and stimulates the quality, accessibility, and affordability of healthcare in the Netherlands (the three pillars of the Dutch healthcare system).Its main legal tasks are to advise the minister on whether care should be included in the basic benefit package of publicly funded health insurance; to distribute public funds among health insurers based on risk equalization; to improve exchange of digital information between healthcare providers; to promote transparency of quality information for citizens; and to stimulate continuous quality improvement of Dutch healthcare (Field document 1).
Our study is concerned with what is called the Institute's quality task although affordability and accessibility also play a role.The Institute improves quality by promoting the development of minimum quality standards.These are publicly accessible documents that describe what is understood as good care from the client or patient's perspective.Examples are performance indicators, clinical guidelines, and other professional standards.Its horizontal forums are representative organizations of patients, healthcare providers (both healthcare organizations and healthcare professionals), and insurers, commonly referred to as the field parties.They are together responsible for developing these standards based on scientific evidence and professional experiences (Field document 2).The Institute maintains a publicly available registry in which these standards are included after procedural assessment.The most important criterium is that all relevant parties are involved in developing and endorsing the quality standard (Field document 3).
When a standard is included in the registry, healthcare providers and other field parties need to comply, although deviations in individual patients are allowed if there is a good reason.The Health and Youth Care Inspectorate (IGJ), another IRA, in turn regulates compliance to the standards (Field document 2).The relevant field parties in a certain healthcare field can submit a standard to the Institute for inclusion in the registry.The Institute can also use its legal instrument-the multi-year-agenda-to set a deadline for certain quality standards to be met by the field parties if it deems quality improvement necessary for public interest in a certain care segment (Field document 3).Another powerful legal instrument-the overriding authority-was created in 2013 to breach the process and guarantee results when negotiations between the field parties about developing quality standards do not lead to consensus (Helderman et al., 2014).Once the development period has expired, the Institute can ask the Quality Council, a board of experts, to develop (part of) the quality standard and submit it to the public registry (Field document 3).Therefore, although regularly the Institute has an executive role and monitors procedures, it occasionally also makes substantive decisions about the content of quality standards.
When the Institute was established, the Dutch government deliberately chose to give this legal overriding authority to a semi-independent agency to guarantee political distance and allow the Institute to develop expertise and gain trust and authority among national healthcare stakeholders.This resulted in a mutual accountability relation since the field parties remain largely in charge of developing quality standards (Field document 4; Helderman et al., 2014).To connect the Institute to the healthcare field, the government decided in 2013 that a board of independent experts (the Quality Council) would advise the Institute in its quality task (Field document 4; Helderman et al., 2014).For example, the Council advises the Institute whether to include a standard in the multi-year-agenda or whether to include a standard in the registry after submission by the parties and procedural assessment by the Institute.However, the Quality Council's role is most publicly visible when the Institute has to ask it to execute the overriding authority and develop (part of) a standard in the public interest when field parties are unable to compromise.At the time of data collection, the Quality Council consisted of 10 members, appointed for 4 years by the Institute because of their relevant expertise in healthcare (Field document 5).During the fieldwork, the board consisted mostly of professors with diverse backgrounds, including medical specialists, hospital directors, economists, and sociologists.Their expertise covered a wide range of relevant healthcare themes, including patient participation, guideline development, public participation, psychiatric care, nursing, primary care, medical specialist care, long-term care, and financing of care (Field notes).In this study, we focused on the development of the national quality standard for emergency care from 2015 until 2020, in which the Institute used its overriding authority.We focus mostly on the time after May 2019 when the Institute and its Council took charge.

Data Collection and Analysis
We used a qualitative research design to study how the Institute dealt with MA and multiple public values in the process.The first author conducted ethnographic fieldwork at the Institute between February 2019 and September 2020.From May 2019 until February 2020, the overriding authority process for emergency care was followed intensively.From February 2019 onward, the first author collected field notes based on 50 hours of observation, including seven meetings of the Quality Council (during which the overriding authority process was observed) and a hearing organized by the Quality Council to consult all the involved national stakeholders (commonly called field parties) was followed.During these observations, the first author supported the members of the Quality Council and Institute employees and helped to organize a conference in June 2019.From February 2019, regular meetings of the Quality Council and Institute were observed.This study was conducted as part a longstanding academic collaboration between the Institute and our research group and was therefore partly funded by the Institute.We shared our findings with policymakers to help them improve their policymaking processes.The close cooperation with the Council and Institute and attendance at closed meetings gave us an in-depth understanding of the overriding authority process.Informal conversations with Institute employees and Council members helped to clarify the meaning of events.Guarantees for scientific independence and critical scrutiny in publishing are written down in a partnership agreement between the Institute and our research faculty.However, doing ethnography always involves striking a balance between being close to the object of research to enable indepth exploration, while not being too close to risk "going native."In our research, after an initial period of gaining entry into the field and relation-building, we enabled critical distance to the field primarily through continuous and collective reflection on our data and interpretations (Moeran, 2009, p. 154).This distance resulted from the use of our theoretical lens and the involvement of the three other authors in the analysis, who, unlike the first author, were not directly involved in the fieldwork.
Supplemental Appendix I provides an overview of the data.In addition to the data collected during meetings and informal conversations, 18 policy documents were used to reconstruct the development process of the standard from 2015 until 2020.These documents included correspondence within the Institute and the Quality Council; correspondence between the Quality Council, the Board of the Institute and the minister; and different versions of the quality standard.We also used four media articles to substantiate politically salient information on field parties' interests.Furthermore, three semistructured interviews with two members of the Council and two employees of the Institute were conducted by the first author.The interviews were recorded and transcribed verbatim.Our respondents member checked the quotes we used.The data (policy documents, media articles, interview transcripts, and field notes) were thematically analyzed by the first author while sensitizing on public values, arguments and perspectives of actors, and accountability dynamics.The analyses were further developed in discussions between the four authors.Our representation of events was member checked by two Institute employees.These conversations revealed details we had not observed such as the development of the standard before the overriding authority process.In the results section, we analyze the development process of the standard.

Results
In this result section, we will give an overview of the 5-year development process of the quality standard.We will zoom in on several phases and events to show how strategic actions of the Institute and its forums resulted in MA practices, how these practices made new public values visible and how this contributed toward consensus between the stakeholders and decision-making by the Institute.In each of the first three sections, we show a different strategic use of accountability practices by the Institute and its forums.In each section, we also show how, because of these practices: (a) particular public value(s) became visible.In the fourth section, we zoom in on the final phase of decision-making by the Institute.

The Start of the Standard's Development: Building Support for Quality Improvement
The Institute used its overriding authority for the first time in 2014 when organizations of insurers, medical specialists, patients, and hospitals could not agree on six quality indicators (optimum volume norms and performance indicators) for emergency care (Moes et al., 2019).During this process, the field parties stressed the need for broader agreements on emergency care (including general practitioners and ambulance services).They agreed that "quality of care in the acute phase is also determined by the extent to which the care chain is functioning" (Field document 7, p. 9).Therefore, the Institute placed the development of the quality standard for emergency care on its multi-year-agenda.At the beginning of 2015, the 11 field parties, namely the representative organizations of ambulance care (AN), medical specialists (FMS), primary care (INEEN), nationwide acute care (LNAZ), Dutch hospitals (NVZ), academic hospitals (NFU), general practitioners (NHG), emergency care practitioners (NVSHA), patients (PF), nurse practitioners (V&VN), and health insurers (ZN) started discussing the current state of quality of emergency care in the Netherlands.National parties were dissatisfied with arrangements within regions, so the Institute organized gatherings with regional stakeholders to map the quality of emergency care chains.They did so using various medical-indication-based patient pathways.According to an Institute employee, the role of the Institute was "facilitating, ensuring that parties come together (..) We first started from the people working at the coalface.(..) That does take more time but if you want to have support you need to take time for that" (employee B Institute).
In this phase, we find that the initial role of the Institute was facilitating and steering the process since the issue was included on its multi-year-agenda.For the first 4 years, the field parties oversaw the development of the standard.Although the agenda provided regulatory pressure, the Institute largely depended on the willingness of the parties and thus had an interest in building relations with them.Horizontal and informal accountability practices directed at learning and gaining trust were dominant for the Institute at this point.Through these practices, the public value quality was made visible and interpreted as the smooth coordination and information exchange in the emergency care chain from a patient perspective who follows a certain care pathway.

Holding the Institute Accountable to Safeguard Public Accessibility and Liveability
In 2018, after 4 years of negotiations, the resulting quality standard consisted of over 100 norms.In addition to agreements on cooperation, coordination, and information exchange, the parties also decided which medical professionals should be available 24/7 at every ED in the Netherlands.The strictness of the norms, specifying the required expertise of the medical specialist in charge of the ED, and of the availability of a professional with geriatric expertise was contested (described in detail later).These norms should enhance the quality of emergency care in the interest of patients.This was deemed particularly relevant for EDs in relatively small hospitals, which are often staffed by young medical specialists with little work experience in emergency care, especially during evening-, night-, and weekend shifts.Parties also discussed the 24/7 availability of geriatric expertise for the ED, which was relevant because the number of elderly people entering the ED is increasing as the population in the Netherlands is aging.Because of multimorbidity his is a vulnerable patient group that requires geriatric expertise (Ellis et al., 2014).
In May 2018, a year before the Institute used its overriding authority on the two contested norms, the involved field parties were considering whether 2 years of relevant clinical experience should be required for the chief medical specialist at the ED.This norm would particularly affect small, general hospitals serving a regional population since, unlike top-clinical and academic hospitals, these hospitals expected difficulty in meeting the stricter personnel requirements due to financial expenses and shortage of qualified workforce.The SAZ, an organization which represents the 28 relatively small general hospitals in the Netherlands, therefore strategically chose to lobby across media (a horizontal forum) which activated the Ministry of Health, Welfare, and Sport (the Institutes' main vertical forum). 1 As an important source of income for hospitals, the SAZ stated that "both the intensive and acute care are a lifeline for hospitals" (Visser, 2018).A spokesperson of the SAZ stated "we need flexible requirements and a little room for arrangement in the hospitals" and warned that stricter norms would "possibly cause hospitals to fall over" (Kiers, 2018).The following quote from an interview in one of the large Dutch newspapers shows that the organization intended to put pressure on the main vertical accountability forum of the Institute, the ministry: "There is a solution to circumvent expensive investments: closing the intensive care unit and the emergency department [interviewer].SAZ: That is possible, but in that case let the ministers of care explain why Dutch citizens in many regions need to travel longer when they need to go to the hospital" (Trouw, 26 May 2018).
This call for the responsible minister to explain regional availability of emergency care was a politically salient issue.Two hospitals had closed in the summer of 2018 because of bankruptcy, which invoked much media attention and societal unrest (Field document 8).Therefore, it was not surprising that the concerns of the SAZ reported in the media evoked political interest (Field document 9; Field document 10).On 19 June 2018, two members of the Second Chamber submitted a motion which was accepted by the majority of the Second Chamber.In the motion the parliamentarians asked the government "to make sure that in the development and assessment of the quality standard for emergency care, a liveability analysis will be conducted" and "to take into account the interests of citizens at the point of maintenance of liveability of the region [regions with general hospitals]" (Field document 9).This liveability could be affected by hospital closures because the resulting unemployment and unavailability of public services may make the region a less attractive place to live.In turn, the minister held the Institute accountable on this issue via a formal letter to its head of the Board of Directors.In this letter, he stated: "I want to ask you to ensure that, in developing the quality standard for emergency care, parties pay sufficient attention to the effects of this quality standard on the quality and accessibility (including the proximity) of the care in regions, including the trade-off between quality and accessibility of care" (Field document 10).
In addition, the ministry had asked the Institute already in 2017 to "ask parties to estimate the financial consequences of a quality standard" and "based on this estimation of parties, ask the Dutch Healthcare Authority (NZa) to conduct a budget impact analysis" (Field document 12).This meant that the Institute had to ask the Dutch Healthcare Authority (NZa, hereafter "the Authority"), another regulatory agency, to conduct a budget impact analysis (BIA) on four norms (including the two contested ones) for which the field parties expected "substantial extra costs."This BIA was published in December 2018.The BIA influenced parties' standpoints on the norm prescribing the required experience of the chief medical specialist at the ED.The BIA showed that the extra costs for realizing this norm in all hospitals were relatively low (1.1 million euro), but it also showed that 52 hospitals (60% of all hospitals in the Netherlands) did not yet meet the norm regarding 24-hour presence (Field document 6, p. 4).This was experienced as problematic since several hospitals expected difficulties in finding the necessary medical staff (Field document 18, p. 36).Despite the preference of most parties for a norm of 2 years and the former consensus, they agreed to compromise in April 2019.By lowering of the norm of 2 years of clinical experience to 1 year, parties agreed that "quality of care will be improved while maintaining affordability and accessibility of care in the whole of the Netherlands" (Field document 13; Field document 14).However, both this norm and the norm regarding geriatric expertise remained contested as we will show in the next sections.
This section shows how horizontal and vertical accountability forums of the Institute strategically interacted which resulted in the Institute being held accountable by the ministry.As a result, the Institute was forced to make sure that parties focused on accessibility of care and livability of the region in the development of the quality standard.These values were thus made visible and were given weight.The strategic accountability dynamics between the Institute's horizontal and vertical forums also appears from how the ministry asked the Institute to let the Authority analyze the budget impact of four norms which influenced the standpoints of the field parties.This accountability practice reinforced the focus on the public value accessibility.

Rendering Account to Horizontal Forums Resulting in a Focus on Quality as Flexibility
In November 2018, after receiving the letter from the minister, the head of the Board of Directors of the Institute asked the Quality Council (horizontal forum) to advise on the broader issue of "how to deal with the assessment of cohesion of care in a region when quality standards are submitted" (Field document 11).In cooperation with the Institute, the Council organized a "dialog conference" on 14 June 2019 called "Good or available care: national quality standards and the consequences for accessibility of care in the region" to deliberate with a diversity of experts and societal stakeholders such as medical professionals, patients, journalists, scientists and executives of healthcare organizations about the tension between nationally applied quality standards (in general) and cohesion and accessibility of regional care.Government officials and representatives of the field parties involved in the development of the quality standard for emergency care were also invited.
The following quote from the head of the Institute shows that it deliberately chose to substantiate its advice to the minister by involving horizontal forums."In the past we might have immediately sent a letter to the minister in reply.Now we first want to know how society feels about the issue.The Institute finds it important to give the minister a broad-based advice.We need to involve parties in the field and cannot solely advise from our office in Diemen.Therefore, the Board of Directors has told the Quality Council to broadly orient itself" (Field notes, 14 June 2019).The participants, the board, and the Quality Council thought the meeting was a great success.The Council collected input for its advice to the Institute and for the overriding authority process which had started in May 2019.We will discuss this process in the next paragraph.When asked about the action of the general hospitals (SAZ) to seek media attention to alert the minister which led to additional effort for the Council and the Institute, a member of the Council answered: "I see that as a necessary thing.So eventually it also has a function.And naturally you [the Council] are in a process [the overriding authority process] and think you are doing well and suddenly something bypasses that you need to take seriously.You can find that very unpleasant, but you also have the right [to decide].So, then it is best to listen carefully to what these people want and why they have this concern and take it into account" (Interview member B Council).The main conclusion of the conference was that experimentation, innovation, and interorganizational cooperation are necessary for healthcare providers to deal with challenges such as shortage of medical personnel.Participants at the conference stated that national quality standards should offer flexibility or "allow room for tailored work" to appreciate differences between regions and not "tie down everything" (Field notes, 14 June 2020).The conference thus made visible and gave weight to flexibility as a prerequisite for public accessibility.
The Institute thus responded to the request of the minister by rendering account to its horizontal forums (the field parties and other societal actors) and involving its Quality Council.Although this required additional effort, it was also seen as a way to collect additional perspectives which led to a focus on flexibility of regulation to safeguard public accessibility and regional availability of care.

Collecting Perspectives on Competency of Emergency Care Personnel as a Prerequisite for Quality
Because the field parties were unable to compromise on two norms on personnel requirements for the ED, the Institute decided to use its overriding authority in May 2019.In a formal letter, the head of the Institute asked the Quality Council "to study the norm regarding the physician at the ED" and "the norm regarding geriatric expertise" and "to establish both norms and submit them to the Institute to enable the inclusion of the standard in the registry" (Field document 15).At this time, the dispute regarding the expertise of the chief medical specialist at the ED had shifted to another issue: how competence should be defined.In March 2019, to the aggravation of the other 10 parties, the medical specialists withdrew their agreement on this norm shortly before the parties submitted the standard to the Institute.Based on progressive insight, the medical specialists preferred a qualitative measure of competence (i.e., entrustable professional activities [EPAs]) over a quantitative measure in years.EPAs are activities that a resident physician needs to be able to conduct before he or she can independently work at the ED (Shorey et al., 2019).The federation of medical specialists (FMS) argued that the duration of work experience does not necessarily mean that a resident physician has acquired the necessary skills to lead an ED.
In August 2019, the Quality Council organized a hearing where it gave 15 parties, including the 11 field parties, the opportunity to express their standpoints.Four smaller parties-the internists (NIV), the geriatricians (NVKG), the general hospitals (SAZ), and the top-clinical hospitals (STZ)-were also invited because they were primarily affected by one of the two norms.At the hearing, the medical specialists expressed their standpoint on the EPA as follows: "The most important argument is that every doctor who works and trains in a hospital knows that there are junior physicians with two years of experience that cannot stand alone at the ED and that there are junior physicians with half a year or nine months of experience who could do it excellent.It is about competences" (FMS hearing, 16 August 2019).However, the other parties were not convinced that EPAs were a reliable measure of competence.They argued that EPAs were too premature and that they had not had enough time to consider them because the medical specialists had only introduced the idea at the very last moment.The organization of emergency care specialists (NVSHA) did not regard EPAs as a good measure of competence and an NVSHA board member voiced his doubts during the hearing: "The proposed EPA and requirements are too minimal, too generic and too non-specific.(..) work experience is a clear criterion.Therefore, we still stand by one year of relevant work experience' (Field notes, NVSHA hearing, 26 August 2019).
At the hearing, the parties acknowledged that EPAs might be a good measure of clinical expertise but argued that the EPAs should first be further developed.Moreover, they stressed the importance of "making a start with the consensus already reached" to improve quality of care for patients.While the medical specialists thought differently about the best way to define clinical expertise, they based their argument on the same public value as the other parties: promoting quality of emergency care for patients.While the parties expressed arguments related to enhancing or safeguarding public values, the partial interests of these organizations also a large role.However, these partial interests were usually not mentioned during the process, although Council members were aware of them.In the quote below, a member of the Council states that the aim of the hearing was to hear the interests separately from the public values.
"It is especially important that we get to hear new perspectives and insights.Who it's from is substantively unimportant.(..) We need to stimulate them and show that we want to get things on the table and that they should not wrap their interests in quality interests.(..) Like 'if we arrange it like this, my influence will decrease.'We do not want to reach consensus; different perspectives can come to the fore" (member Quality Council, Field notes, 1 July 2019).An example of a possible underlying interest is competition between different medical disciplines on their position in the hospital.When asked about this in a media interview on the quality standard, a spokesperson for medical specialists answered: "that could surely play a role, we are after all professional interest groups."However, in the rest of his answer he called upon a public value: "but eventually it is about the patient: the patient deserves the best quality of care" (Maassen, 2019).
For the geriatric expertise norm, parties agreed that in an ideal situation geriatric expertise at the level of a medical specialist should be available at the ED, by telephone consultation within half an hour or in person within 2 hours.However, because of extra costs and the shortage of geriatricians and internists with geriatric expertise, parties saw this as unfeasible.Therefore, they agreed to include a geriatrician as a recommendation in the standard and reached consensus at a norm on the timely availability of geriatric expertise on the level of either a specialized nurse practitioner or a medical specialist.In this case, the nurse practitioner could still consult a medical specialist but not necessarily a specialist with geriatric expertise.The spokesperson of the general hospitals (NVZ) explained this at the hearing as follows: "The deployment of a nurse practitioner with a geriatrician as backup is the ideal situation that we would all prefer [all parties].(..) The current formulation as stated in the standard ensures that geriatric expertise is available for elderly patients at the ED.(..) We consider this formulation financially and organizationally feasible and it provides hospitals-small, big, average-with more flexibility to provide the necessary expertise.This is important because a hospital has to implement this, and the inspectorate will monitor it" (NVZ Field notes, 26 August 2019).
However, the organization of medical specialists (FMS) and two of its daughter organizations, the organization of geriatricians (NVKG) and the organization of internists (NIV), opted to include a geriatrician as the norm. 2 They argued that elderly patients entering the ED are very vulnerable and specific expertise is needed to ensure they receive quality treatment.They also argued that providing a geriatrician would be feasible; new geriatricians are being trained, so would be available and the increased efficiency would save money, covering the extra costs: "These are really the most complex patients there are for whom you want to call upon the best expertise there is.There were arguments stating 'is that feasible?' [expressed by other parties].The expectation is that with the number of geriatricians being educated it should be feasible within a few years regarding availability of personnel.(..) The costs to deploy a geriatrician early on yields that less people need to be hospitalized and that it eventually leads to efficient care where the costs really will be recouped" (NVKG Field notes, 26 August 2019).These three parties might have had specific interests in this norm, which were not made explicit in the process.For example, this norm could affect the position of geriatricians in the hospital relative to other medical professionals.
This analysis showed how, through the strategic action of the medical specialists (FMS) to not endorse the two norms, the debate again shifted to another issue related to the value "quality," that is, how to define clinical competence.Partial interests seem to have played a role in decision of the medical specialists to undertake this action.However, these partial interests largely remained implicit; parties called upon public values of accessibility and quality of care instead of expressing these interests.The action led to the overriding authority process in which the Quality Council (a horizontal forum of the Institute) initiated more horizontal accountability practices.It tried to collect perspectives in the hearing and rendered account to additional forums, involving additional parties (the internists and the geriatricians).

The Final Decision: Finding a Balance Between Quality and Accessibility
After the hearing, in August 2019, the Quality Council gave the parties the opportunity to submit their final standpoint on paper.Independent (emergency care) experts, who were also present at the hearing, were also asked to submit their advice to the Council.Before the hearing, the Council did not want to "think in solutions" yet but preferred to first thoroughly explore "the possible underlying issues at stake" (Field notes, 6 June 2019).The Council collected input from experts and field parties, then started to think of possible solutions and specific formulations for the two norms and scrutinized the whole standard.At the request of the Council, the Institute hired an external expert to assist the Council with this.The Council found this important because, as a former emergency care physician, this expert is "independent of the field parties but speaks the language of the field" (Field notes, 16 June 2019).To advise the Institute in the overriding authority process, the Council also formulated its own six starting points based on "the public perspective" such as "right care in the right place" and "taking the accessibility into account" (Field document 18).This societal perspective of the Council is experienced as valuable by the Institute.In addition, the value of the Council's prominent role in the overriding authority process lies in its ability to incorporate valuable perspectives.Because its members work in healthcare practice but are independent of the field parties, they possess relevant expertise as an Institute employee explains: "I think that this connection with the field is very important for the Institute.The Council has that function that there is more connection with the field than we have ourselves.And in that way the voices of the field also reach us in a different way than via the representative organizations which are our usual counterparts" (employee A Institute).
In its final decision on the norm prescribing the required expertise of the medical specialist supervising the ED 24/7, the Council chose for a norm of 1 year of relevant clinical experience.This decision agreed with the standpoint of most field parties, except that of the medical specialists.However, the Council also stated in the final standard that "years of experience is only a limited measure for the assessment of competence.Entrustable Professional Activities (EPA's) are developed for educational purposes and seem to be promising instruments to better assess competence, also in emergency care" (Field document 16).Therefore, the Council gave the medical specialists (FMS) and the organization of general hospitals (NVZ) the task to experiment with EPAs in emergency care for the next 2 years, in close cooperation with the other parties, with the view of adapting the norm later on based on these experiments.The Council also wrote 15 other recommendations for redevelopment and implementation in the standard (Field document 16, p. 55).
Concerning geriatric care in the ED, the Council went against the norm formulated by the field parties and chose to prescribe that a geriatrician or internist-geriatric specialist should be available 24/7 for telephone consultation and to see the patient for diagnosis or treatment whenever necessary.The Council found 30 minutes to wait for a telephone consultation and 2 hours to wait for an in-person consultation to be too long so decided to abandon these limits: X: "For a small hospital it remains difficult to have emergency care facilities in the house.They are then placed in an inferior corner."Y: "It is also about the room that you provide for the situational."X: "In addition, 2 hours until geriatric expertise is available is very long and do you need to formulate it so specifically?"(Field notes, 29 August 2019).
On the one hand, the norm became stricter.On the other hand, the Council wrote that "the availability can also be regionally organized" (Field document 16), giving healthcare providers the flexibility to find innovative solutions, such as sharing personnel.On 6 December 2019, the Council sent the draft version of the quality standard to the parties.The Council is legally obliged to consult the parties and did so in a consultation round.The parties were asked to respond to the quality standard and the Council dealt with the feedback in an "accountability document."The Council stated in reply to some of the comments that it would add some specifications to the standard.However, it did not alter the two norms, even though the organization of academic hospitals (NFU) had stated that it preferred a stricter norm for experience of the ED specialist.The Council replied: "The Quality Council believes that with the current formulation a good balance between quality and accessibility of care is found" (Field document 17).In the consultation round, several parties expressed their appreciation for how the Quality Council carefully handled the process.The organization of primary care stated to be pleased because the Council had taken "a weighted decision so that the standard can be included in the registry of the National Health Care Institute."When the whole standard was finalized and the whole Council agreed, the Council submitted the standard for inclusion in the Institute's registry.In February 2020, the Institute endorsed the quality standard and included it in its registry, making the standard legally binding for all healthcare providers, professionals, and insurers.
The overriding authority process shows the complexity of the horizontal accountability relations of the Institute.The Institute is legally obliged to ask the Quality Council to oversee the overriding authority process.In turn, the Council must consult the field parties in the process.Besides this, the Council also made additional efforts such as involving experts and carefully scrutinizing the whole standard.Because of this multiplicity of accountability practices the overriding authority process took 1 year which was relatively long.However, it also enabled the Institute to incorporate multiple public values in the development process of the standard which increased societal support for its decisions.

Discussion
Our empirical study shows that independent agencies often operate in dense and complex accountability networks.We show that, despite its position as a regulatory agency, the Institute depends on building a good relationship with its regulatees and other horizontal forums.Our study touches thus upon the ongoing discussion about the width of the concept public accountability that has been broadened particularly to refer to forms of horizontal or social accountability (Brummel, 2021;Willems & Van Dooren, 2012).Like previous studies, we argue that accountability practices are manifold and can be mandatory, voluntary, mutual, learning-oriented, politicized, and strategic (e.g., Busuioc & Lodge, 2015;Maggetti & Papadopoulos, 2018;Scott, 2000).
We contribute to theory about the benefits and drawbacks of the multiplicity of these practices.Literature on MA has stressed several drawbacks such as inefficiency and complexity of decision-making and a high workload for policymakers (Cohen et al., 1972;Flinders, 2011;Schillemans, 2010Schillemans, , 2011)).We also observed these drawbacks in our study.The development of the quality standard was a complex process that involved many actors and took 5 years.Accountability practices caused delay and administrative burden.These drawbacks were reinforced because the Institute started from a regional bottom-up approach and the Council carefully analyzed the whole quality standard.On the one hand, MA was necessary since it allowed the Institute to render account through creating coherence out of seemingly conflicting demands and values in a continues process.New accountability practices such as the dialog conference and the hearing made other values visible.However, accountability practices not always added something new.At several times, a process was created while consensus was already reached.This happened for example when the Quality Council focused on parts of the quality standard that parties had already agreed upon, delaying the Institute's decision on the two disputed norms with the risk of causing irritation among the parties involved.
In contrast to existing literature on the benefits and drawbacks of MA, we argue that the inefficiency and complexity of MA do not have to be drawbacks.Accountability and efficiency do not always have to be conflicting values as de Graaf et al. (2016) show.In our case, on the one hand, MA served as a way to deal with conflicting values and on the other hand, to render account by creating coherence out of conflicting values.This enabled the field parties to reach consensus and the Institute to make decisions.MA became redundant in a negative sense when accountability resulted in a timeconsuming process in which uncontested issues were questioned.Because of this, efficient decision-making and implementation in the public interest were impeded.This is particularly troubling since promoting efficiency is a common motivation for governments to delegate tasks to semi-independent agencies.This was also why the overriding authority was created and attributed to the Institute.We recommend to policymakers of agencies to be aware of this possible risk of redundant MA, but that they do not shy away from MA as it can fruitfully deal with conflicting values to enhance efficiency of decisionmaking and societal support for these decisions.
Our detailed case study on MA and public values contributes to theories about the widely studied policy challenges of dealing with conflicting values and public accountability.Conceptualizations of values are often used interchangeably with the concept of interests.For example, Schillemans (2010) speaks of values as budgetary discipline, improving operations, and stability of the organization.In the widely used definition of Thatcher and Rein (2004), values are defined as the "ultimate ends of public policy."This latter definition acknowledges that public values are distinct from partial interests because they concern collective public ends.In our case, it was not necessary to discuss partial interests or to sort out how values were related to interests.The focus on public values was a way to set partial interests aside and to focus the discussion on the general public.We saw that actors formulated their interests in terms of different interpretations of the three public values: quality, accessibility, and affordability of emergency care.Following Gallie (1956) and Eriksen (2021), our case shows that values are essentially contested concepts.Our analysis thus contributes to conceptualizations of values by showing that, although driven by interests, value discussions move beyond them.
Our case fits the sociological idea of decision-making as a continuous process rather than of separate phases such as assessment and appraisal (Kleinhout-Vliek et al., 2020).Like Lindblom (1959), our case shows that goals were not uncontested and clearly predefined but created during the process.The fluidity of preferences, processes, and participants (Cohen et al., 1972) enabled a process in which values were constantly made visible and given weight.This continuous process is different from less dynamic institutional strategies for dealing with value conflict such as cycling and firewalls (Thatcher & Rein, 2004).Also, we showed that not only the Institute and its Quality Council, but also other actors took part in the weighting of public values.When policymakers are aware of these differences, they can remain open to values that become visible at different moments and are made visible by different participants.We recognize that this fluidity, multiplicity, and overlap of accountability forums and practices also blurs who is accountable for what and to whom as our case also shows.On the one hand, clear agreements about roles and responsibilities of actors in decision-making processes can prevent dysfunctionalities (Koppell, 2005).On the other hand, too solid agreements and fixed roles restrict the flexibility that allows incorporating multiple values.Further empirical research on accountability dynamics could provide insight in how to deal with this tension.
Our study has several limitations.Developing the quality standard took 5 years and the standard consists of more than 100 norms.Therefore, we recognize that we have not given a complete overview of the entire process in our paper.However, the two contested norms were most relevant for our study since the observed accountability practices and value conflicts mainly concerned these norms.We concentrated in depth on several relevant events such as the overriding authority process on the two contested norms and the lobby of the general hospitals which invoked political interest.Since we did not conduct interviews with representatives of the field parties, we could not consider discussions within field parties.However, since we concentrated on specific events and focused on the role of the Institute as an independent agency, this did not complicate our analysis.Further research could study how the accountability forums of agencies (such as the ministry and the field parties) perceive MA and the role of conflicting values.Finally, although our case study of decision-making by a single agency and the use of ethnography enabled us to give a detailed overview of the accountability dynamics involved in the process, we recognize the limits to the generalizability of our findings.Particularities of our case, such as the influential position of the Quality Council as the Institute's advisory committee and the strong consociational tradition of (health) policymaking in the Netherlands influence our findings.Therefore, further research on the role of MA and conflicting values in other types of independent agencies with different tasks and in different countries, will provide further insights into the generalizability of our findings.

Conclusion
In this paper, we explored in detail the incorporation of values as a possible benefit of MA.We show how the Institute strategically dealt with both the challenges of MA and conflicting values by using multiple accountability practices to create coherence out of conflicting values.We showed how MA brought different interpretations of public values into the process which enabled the involved actors to reach consensus and the Institute to take decisions.First, we showed that MA was constituted by the involved actors using additional accountability practices and forums such as the media.These practices supplemented the regular and legally established practices which resulted in a multiplicity of accountability practices.Second, we showed how, as a result, a process was set in motion in which values were continuously made visible and given weight by the involved actors.In addition, we found that, in our case, the discussion about public values was a fruitful way to move beyond a discussion about partial interests because it focused the discussion on public interest.While the field parties deployed accountability practices out of partial interest, they did not explicitly voice these interests.Instead, the field parties used a public value, such as regional availability of emergency care, to substantiate their standpoints.This does not mean that partial interests were not at stake, but they largely remained implicit.Finally, in our case, MA also impeded efficiency since it increased the complexity of the process and caused delay.To conclude, we argue that finding a balance between MA and efficiency is an important challenge from a public perspective.Determining when accountability practices are necessary and constructive and when they become redundant in a negative sense is a complex challenge.Further empirical research on the role of accountability and public values in decision-making processes of agencies can provide further insights into how to deal with this challenge.