The theoretical and practical dilemma of creating a new organisational design for a municipality using sub-municipal units

Organisational design is an important factor in determining the survival of organisations . The organisation and design of municipalities and local government is no exception to that rule. Concluding an amalgamation process indicates that a new organisational design must be introduced. The rule of thumb is that such proposals do not deviate much from the organisational tradition of local government in country or region in question. However, there are exceptions to that rule. Following an amalgamation of four Icelandic municipalities in 2020 a new organisational design was introduced. Consequently, a municipality with sub-municipal units was established for the first time in Iceland. Sub-municipal councils with a mixture of directly and indirectly elected members were created in each of the previous four separate municipalities, creating a second administrative tier within the municipality. The aim of the new design was to address problems inherent in amalgamation processes such as alienation of smaller communities within the new municipality often resulting in feelings of low levels of democratic legitimacy in many newly established municipalities. Another aim was to ensure that the municipal administration and services were organised aligned with local needs. The study aims to shed light on both the theoretical and practical dilemma of creating a new organisational design for a new municipality by using an interdisciplinary approach combining organisational theory with theory and studies on local government. Therefore, the study


Introduction
Decentralisation has become somewhat of a buzzword in past decades, with supranational and national governments heavily advocating decentralisation as both a democratic and an efficient way to organise governance (Hlynsdóttir 2020). This trend has been noticed at the Icelandic local level, as in the past three decades, new tasks and responsibilities have been systematically decentralised to the local government. The Icelandic local government system is synchronised; all local authorities are, by law, supposed to provide their citizens with the same types of services, irrespective of size. Currently, the municipal size ranges between 40 and 136,000 citizens, with around half of all municipalities having less than 1,000 inhabitants. Consequently, the heavy emphasis on task decentralisation has led to a flood of largely voluntary amalgamations in recent decades. The number of local governments in Iceland has dropped from 202 in 1990 to 64 in 2022-by 68%. These territorial changes have frequently led to problems for many newly established municipalities as they struggle to provide services to a vast territory, often with a low population density. Concurrently, this has also led to many peripheral communities feeling left out and alienated within the new municipality, leading to low levels of democratic legitimacy in many newly established municipalities. To tackle this problem, a new type of municipality with sub-municipal units was established for the first time in Iceland in October 2020. A large municipality in the rural east of Iceland was amalgamated with three very small municipalities. In each of the four original municipalities, a sub-municipal council was established with a mixture of directly and indirectly elected members. These councils were given responsibilities in the field of local planning, as well as in issues directly concerning individual communities, such as opening hours of recreational facilities. In general, the local government in Iceland is organised into one tier, with all local councils legally responsible for the same type of tasks. However, a special article was added to the local government act (Art. 132 no. 138/2011) in 2021, giving local authorities in the new municipality Múlaþing (Mulathing) leeway to test a new type of democratic organisation organizational design of local government. The act states that the experiment must take place for at least eight years. The Múlaþing experiment will therefore be in place between 2020 and 2028. The administrative organisation of Múlaþing serves as a prototype for a possible new design of local government. If successful, it is more than likely that this will be implemented more widely at the Icelandic local level. Hence, it is an experiment with a clear beginning and an end.
This paper explores the design of the new municipality from the perspective of organisational design in addition to research on municipal reform. This clear devolution of power within a single municipality is an interesting experiment in an otherwise closed and synchronised system.

STJÓRNMÁL & STJÓRNSÝSLA
This paper explores the following question: How successful is the Múlaþing experiment so far in relation to the aims of the new organisational design?

Theoretical background
Public management reforms have been high on the agenda for a long time. Whether it is rational hierarchical planning of the 1960s and 1970s, new public management trends at the end of the 20 th century or the more obscure trend of governance in the past two decades promise to improve public management (Pollitt & Bouckaert 2011). Local governments have not been excluded from this trend. Local government reforms have been on the agenda throughout Europe over the past few decades. The main trend has been on decentralisation, where the focus is on moving tasks and authority from state government onto lower tiers of government-in most cases, local government. There have also been other experiments, such as removing tasks from local government re-centralisation or moving tasks back onto local-level re-municipalisation. In relation to these functional reforms, extensive territorial reforms have taken place, inspired by emphasising participatory democracy and bringing decision-making closer to the citizens. Simultaneously, the internal structure of municipalities or cities has been tested. In contrast to the issue of work division between higher and lower tiers of government or the territorial boundaries of local authorities, much less has been written on these internal structural changes in municipalities-the work of Hlepas et al. (2018) Sub-municipal governance in Europe. Decentralization beyond the municipal tier is a notable exception. The concept of sub-municipal entities stands for the creation of units of government within a municipality that go beyond the normal structure of, for example, the councilcommittee system.
There are many kinds of sub-municipal entities within municipal boundaries, and in many cases, these entities rely on the participation of societal actors or civic society. In such cases, the focus of study is on the relations between these civic actors and the local government (see e.g., Heinelt et al. 2021). However, the presence of such societal or civic actors does not change the organisational design of the municipality. As a point of departure, this paper uses Hlepas et al.'s (2018) definition of sup-municipal units: '[M]ultipurpose sub-municipal units with territorial competence and democratic legitimacy which do not constitute a fully independent layer of local government and do not possess exclusive territorial jurisdiction over their local affairs and citizens ' (p. 20). Sub-municipal units also vary based on the level of population density; hence, cities tend to use sub-municipal entities to give voice to different parts or neighbourhoods of the city. In rural areas, sub-municipal entities are more likely to represent a well-defined territorial area based on tradition and frequently have deep historical roots. In such cases, sub-municipal entities are thought to safeguard local identities and communities (Swianiewicz et al. 2022). The creation of sub-municipal units of this type often follows extensive amalgamation processes. Consequently, some have argued that such sub-municipal entities are a specific 'form of compensation for the loss of autonomy and self-regulation' (Kersting & Kuhlmann 2018, p. 128). Based on this argument, the

STJÓRNMÁL & STJÓRNSÝSLA
The theoretical and practical dilemma of creating a new organisational design for a municipality using sub-municipal units sub-municipal unit has no other purpose than to smooth the transition from a system of many small, community-based municipalities to a larger municipality. Studies on submunicipal designs have shown that there is a real danger of a power struggle between the mother council and the sub-municipal councils (Hlepas 2018). Decentralisation has long been seen as a remedy for various problems within the organisation of the modern state (Conyers 1983). Supranational agencies and international treaties are geared towards it, vehemently arguing for its virtue. Schakel (2021) identifies two logics and four drivers behind the decentralisation trend. Functional logic suggests that there is an optimal scale for services. Hence, Europeanisation and the expansion of the welfare state are drivers behind decentralisation. The latter is the identity logic driven by territorial identity and democratisation. There are several ways of conceptualising decentralisation, but it is common to separate fiscal, political and administrative decentralisation (Dardanelli 2021). Decentralisation has many practical implications; however, the fundamental argument for it is the strong notion of the democratic virtues of the 'local'. Purcell (2006) has conceptualised this tendency in the concept of the "local trap". In short, this means that scales based on the local are thought to be better than other scales. Consequently, we tend to believe that countries that have not extensively shared tasks and authorities with tiers of local governments are less democratic than countries that have. In this line of argument, Sweden is more democratic than France is. Therefore, Purcell (2006) argues that the local is frequently conflated with democratisation and again the more local government tiers or sub-municipal units there are, the more democratic the society must be. Consequently, as democracy is good for society, there is more social justice in the wider society. Local communities also tend to be conflated with participatory democracy based on the assumption that more 'local' decision-making will automatically lead to more popular participation. The main point of Purcell's argument is that there is nothing inherent in the locality; it may or may not bring better and more democratic decision-making or more social justice to society. Jun and Musso (2013) also pointed out the danger of the 'not in my backyard' phenomenon (NIMBYism) in geographically determined communities. Drew (2020) made the same line of argument when he points out that amalgamations do not automatically fix problems, such as lack of professional capacity, service quality or financial instability. Nevertheless, it is common for governments at all levels to view amalgamations as a remedy to various problems. As more tasks become decentralised to the local level, the external and internal pressure on smaller municipalities increases to amalgamate into larger municipal entities to be more efficient and effective (Swianiewicz et al. 2022).
This raised questions about the degree of intentionality. How much can we actually predict the outcome of our actions when it comes to designing a new municipal structure? Swianiewicz et al. (2022) argue that there are no 'God-Like designers' (p. 10) out there with the perfect layout for a local government. Municipalities are multipurpose organisations with multiple and often conflicting tasks. They are also some of the longest-lasting public institutions-many have strong historical roots, reaching hundreds of years. Although the modern version of the internal democratic structure STJÓRNMÁL & STJÓRNSÝSLA of most municipalities has its origins in the 19 th century, their territorial identity is often much older. Thus, amalgamation is not only disruptive to the internal organisation of municipalities but also undermines place and community identity. Designing a new municipality structure is, therefore, no small task. The modern municipality must be both democratic and efficient while staying true to its community roots. Organisational design is a key factor in determining the survival of an organisation, and municipalities are no exception to that rule. A classical definition of the organisation concept stems from March and Simon (1958, p. 2). 'Organizations are systems of coordinated action among individuals and groups whose preferences, information, interests or knowledge differ.' This definition invokes the idea of rationality, in which an organisational structure is designed in a logical and systematic way, normally phrased as bureaucracy. Despite new public management emphasis on market incentives and new ways of organising, modern government still relies on bureaucratic formalities, such as hierarchy (or structure), rules, policies, procedures and official recordkeeping, thus demonstrating a high level of adaptability of the bureaucratic structure (Huczynski & Buchanan 2017).
Advocates of contingency theory, such as Galbraith et al. (2001), argue that there is no ideal way to design organisations. Different organisations need different designs based on circumstances that may be new and special for the organisation in question. The foundation for organisational design is fundamentally twofold: differentiation and integration (Lawrence & Lorsch 1967). Differentiation refers to the structure or division of labour in organisations and how organisational goals are broken down into tasks, which are then allocated to different sub-units or individual members of the organisation. Integration is a synonym for both cooperation and coordination between individuals and sub-units within an organisation. Burton et al. (2021) point out that a new design is more than just drawing a new organisational chart. How differentiation and integration are organised and balanced will have a profound influence on the success of the organisation. Mintzberg's (1983) classical work on organisational design identifies five basic types of organisational structure: simple structure, machine bureaucracy, professional bureaucracy, divisional form and adhocracy. Each type is a configuration of different parts and sets of coordination mechanisms. The typology has proven to be a powerful analytical tool in organisational design.
The traditional bureaucratic organisational design has proven remarkably stable. Huczynski and Buchanan (2017) demonstrate, however, that although a functional or divisional structure is still a popular design, other variants, such as the matrix or teambased structure, are becoming no less popular. More recently, they have shown how organisations are moving beyond traditional bureaucratic features in their organisational designs. Public institutions, on the other hand, tend to be more stable and less prone to fundamental changes in their organisational structures. Nevertheless, public institutions frequently fall under the spell of fashion demonstrated in the creative design of organisational charts, and local governments are no exception to that rule.
As demonstrated by, for example, by Svara (1990), there may be differences in the level of separation and overlapping between politics and administration at the local level 6

STJÓRNMÁL & STJÓRNSÝSLA
The theoretical and practical dilemma of creating a new organisational design for a municipality using sub-municipal units based on place in time, number of tasks or forms of local government. Thus, by examining the organisational charts of a local government, it is possible to identify special nuances to the local government in question. Burton et al. (2021) pointed out that it is important to take a holistic approach to examining organisational charts. This includes examining both structural and human elements. This means examining such diverse items, such as goal strategy and structure, as well as more human elements, such as tasks and agents. This approach also includes exploring the level of coordination and how control and incentive mechanisms are built into organisational design.

Local Government and the Case of Múlaþing
In October 2018, the local councils of four municipalities in the eastern part of Iceland-Fljótsdalshérað (population: 3,600), Seyðisfjarðarkaupstaður (population: 685), Djúpavogshreppur (population: 472) and Borgarfjarðarhreppur (population: 109) -decided to formally begin negotiating a merger. One year later, in October 2019, a public referendum took place where a merger was agreed upon, and a new council was elected in special local elections in October 2020 (Þórhallsson & Reynisson 2020). The new municipality was named Múlaþing; it covers 10.671 km 2 (of which large parts are unhabitable), with 5,020 inhabitants. The municipality is the second largest in size and 11 th in population (Statistics Iceland 2022). The largest urban centre is Egilsstaðir, with a population of around 3,000. From the urban centre of Egilsstaðir, there is between an hour and hour-and-half drive to each of the other three villages. However, driving between one end of the municipality and the other may take between 2 and 2.5 hours, depending on road and weather conditions. To reach each village, one must pass mountainous roads that may become unpassable in the winter. Thus, the municipality of Múlaþing is an exemplary case to demonstrate problems facing local councils in a newly merged municipality with long distances between its different corners. Múlaþing is also an example of a new design for an Icelandic municipality. There are two tiers of government in Iceland: state and local. Each municipality, irrespective of size, has the same responsibilities towards its citizens (Valsson 2014). Icelandic municipalities also have an extensive number of tasks and authorities, similar to other Nordic municipalities. This has created heavy restraint on many municipalities, frequently hasting in the process of amalgamation, as many small municipalities are unable to provide the level of services required by the state. The number of tasks at the local level has also led to an elaborate web of intermunicipal cooperation schemes, with municipalities trying to compensate for the lack of an economy of size (Hlynsdóttir 2018). In Iceland, a referendum is legally required to take place if an amalgamation is planned; therefore, the state has a limited scope for top-down planned amalgamations. Although a large number of amalgamations have taken place, there are also well-known examples of such schemes failing-in most cases, due to the unwillingness of smaller municipalities to amalgamate with larger neighbours. The main reason for this negative view of amalgamations has been the fear of smaller communities being alienated or left powerless in a larger municipality (Hlynsdóttir 2018).

Formal arrangements of the local government in Iceland
Municipal councils are usually proportionally elected, except in very small municipalities, using bloc voting (personal voting without parties) (Hlynsdóttir & Önnudóttir 2022). Municipalities with 2,000 citizens or less may use either five-or seven-member councils, municipalities with 2,000-9,999 citizens must use 7-11-member councils and larger municipalities can choose between councils' size of 11-31-member councils, based on the population size. The Icelandic system is a council-committee system (Larsen 2005), with the municipal council playing a central role. Formally, all authority lies within the council. It is presided over by the council leader, who, apart from the responsibility of organising the council meeting, does not have formal rights. Formally, the council is the central decision-maker of the local authority. All decisions must be made by the council or based on the clear devolution of the council's power. The municipal council may discuss any matter it believes may concern the municipality.
The municipal council may decide to set up an executive committee; however, councils with five members are not permitted to do so. The executive committee is formally one of the standing committees; it does, however, have considerable authority, as it is responsible for the day-to-day management of the municipality in collaboration with the municipal chief executive. The council may choose to provide the executive committee with considerably more authority and power than stated by the Local Government Act.
The Local Government Act does not state which standing committees should be in place. This is derived indirectly from other acts, such as education, planning or social services. However, the council may decide to cut down the number of committees by merging them or taking over their tasks. Hence, the number of committees and comparability are difficult to assess; however, standing committees are usually displayed clearly in the organisational chart. There are many other committees at the local level, both internally and as part of intermunicipal schemes. In many cases, these committees or arrangements do not appear on organisational charts. Due to differences in size, the internal organisation of municipalities varies considerably between municipalities. Reykjavík City has, without any doubt, the most complex administrative setting in Iceland; nevertheless, all municipalities are organised in a very similar manner, with a complex mixture of political and administrative lines of authority. This older version of the Reykjavík chart of organisation, shown in Figure 1, presents a very traditional setup of the administrative and political apex in an Icelandic municipality. The broken line depicts the political line of authority, with the city council at the top and the executive committee right below it. Each standing committee is then placed in the administrative department in question. Political committees do not formally have any authority over day-to-day management; however, they work very closely with the administrative unit in question. This informal authority is not clearly visible in the chart. In this case, the position of chief executive is occupied by a member of the council. Hence, the mayor of Reykjavík is an executive mayor with all the responsibilities of a chief executive.

STJÓRNMÁL & STJÓRNSÝSLA
The theoretical and practical dilemma of creating a new organisational design for a municipality using sub-municipal units Before Múlaþing municipality was created in 2020, each of the four municipalities behind it had its own internal organisations. However, only the largest municipality Fljótsdalshérað had an elaborate version of an organisation chart. As seen in Figure 2, in this case, the organisational chart separates between the political and administrative arms of the local government, which is a rather unusual setup for an organisational chart because most of them use a simpler variation of the chart, as shown in Figure 1. In Figure 2, there are no lines of authority demonstrating a chain of command; hence, considerable background knowledge is necessary to understand the functional relationships between different units. The municipal council is on top with the executive committee on the left side and intermunicipal units on the right side. Below are three departments with the political standing committees in question, written in boxes. The chief executive is on top of the lower level of the chart. In this case, the chief executive is hired from outside the council; hence, a politician is not a direct supervisor of the administration, as is the case in the city of Reykjavík. The chart uses colours in place of lines to show the connections

Method
The Múlaþing experiment is due to last a minimum of eight years from 2020-2028. The first election took place in October 2020, and the second was held in May 2022. Thus, this study only looks at the first 21 months of the new municipality; its main purpose is to create a point of departure for future investigations of the case of Múlaþing.

Designing a new municipal structure
The preparations for the new municipal arrangements began in October 2018, as the four municipalities agreed to begin negotiations. A referendum took place on 26 th of October 2019 with majority of citizens in all four municipalities agreeing to amalgamate into a larger unit. A special election took place in October 2020, with a new council being elected in the new municipality Múlaþing. In May 2022, the second elections in the municipality were held, as they became part of the normal four-year election cycle of local government councils. A lot of energy and work was put into the preparation leading up to the amalgamation referendum in 2019. In addition to the aforementioned reports, community meetings were held in all municipalities, and plans for this new type of municipality were introduced in local and national press and venues.
In a speech at the Fiscal Conference of the Federation of Local Authorities from October 3 rd 2019, Jóhannesson drew out that 'the main aim of the amalgamation process was for the amalgamation to lead to better services, stronger administration and more success in regional and transportation issues'. However, at the same time, Jóhannes son pointed out that a new type of organisational structure was 'a vital key to influence in less populated communities in a new municipality'. At the heart of the plan

STJÓRNMÁL & STJÓRNSÝSLA
The theoretical and practical dilemma of creating a new organisational design for a municipality using sub-municipal units was the idea that it was vital to preserve the "core of each community" and its "uniqueness and local empathy". Hence, in the reports of the preparation committee, as well as in other published materials, this very strong approach to localism shines through. The main idea behind the proposed organisational structure was to ensure that locals had direct influence and formal authority over localised issues. Thus, slogans such as 'real decentralization, not just consultation' were used in the phase leading up to the referendum. At the same time, the aim was to keep the administration simple and effective but with a strong connection to all parts of the new municipality. To succeed, the preparation committee suggested utilising e-government more than before, for example, by using online meetings for committee meetings. Additionally, there was a plan to make the municipal website more interactive in such a way that applications for local services could be made online. Furthermore, there was a special focus on setting up a local service desk in all four areas of the new municipality. Finally, the amalgamation would strengthen specialisation within the municipality and create grounds for more effective administrative operation within it. Jóhannesson concluded his speech by saying that the influence of smaller localities would need to be supported in particular, and this could be done in addition to the ideas mentioned above by providing local elected councillors with satisfactory compensation for their work. The key point was, however, that in addition to the understanding of the municipal administration of local circumstances, the people living there would need to participate in politics and civic life. As pointed out before, the Act of Local Government (Art. 132 no. 138/2011) needed to be changed in parliament to allow the devolution of authority from the municipal council to the individual home rule councils. The resolutions of each home rule council show that in relation to giving the home rule councils "real" authority, they were given specific rights in decision-making in relation to land use and shepherding, culture topics, such as culture houses and camping areas, and opening hours of restaurants, to name but few. Most importantly, they were given direct and final authority over specific areas within local planning (including harbour planning) and the evaluation of the environmental consequences of development.

Answers to the closed survey questions
The respondents were asked to rate the workload of the following tasks (based on task description of the home rule council) in the work of the home rule councils on a five-item Likert scale: local planning, environmental issues, permits, agricultural issues, culture, commenting on service tariffs and other undefined issues. The task of local planning was by far the most important, as 70% of the respondents believed it was a high or very high part of the workload. However, other undefined issues were also very high, with around 37% of the respondents stating it as having high importance. Home rule councils may take up any issue they believe is important to their communities. The municipal council may also delegate tasks to the home rule council that are not stated in the public resolution. This response indicates that there are issues not stated in the resolutions that are taking up considerable time within the home rule councils.

STJÓRNMÁL & STJÓRNSÝSLA
As pointed out earlier, there are several underlying aims for amalgamation. The respondents were asked (on a five-item Likert scale) whether they believed these aims had been fulfilled. In sum, the respondents were overall happy or very happy with the results in relation to preserving the core of each community, its identity and influence over the local service provision. Two questions revoked more negative responses. For the question "Do we minimise the control of the largest community in Múlaþing?" 27% of the respondents disagreed, and regarding the question "Does the new municipal organisation ensure a simple and efficient administration?", 36% of the respondents disagreed; only 14% agreed with it. This last statement refers to one of the most important aims of the preparation committee. It must be stated that 71% of the respondents believed that the administration had a strong connection to each local community. Hence, the connection between the central administration and individual localities is in place, but the answers raise questions in relation to the effectiveness and efficiency of the organisational design of Múlaþing.
The respondents were also asked whether they thought the municipal services had improved since the amalgamation. Overall, the respondents believed that the service provision had improved much or very much in all the smaller communities: 71% in Borgarfjörður eystri, 75% in Djúpivogur and 88% in Seyðisfjörður. However, 62.5% of the respondents believed that the level of service was the same as before in Fljótsdalshérað. The responses suggested that the service level needed to be improved in the smaller communities, while amalgamation did little for the better or worse in the largest community.

Answers to open-ended survey questions
The responses frequently showed a strong belief that the organisational design strengthened smaller communities and prevented alienation.
The pros are that local citizens are able to influence specific issues and get local issues into the system. That I find is the most important asset, on the other hand I believe that the home rule council should be unpolitical and represent the majority of citizen rather than the majority in the municipal council. That the home rule council speaks for the citizens first and foremost. That has not always been the case.
This individual sees the arrangement as a positive development. However, in very small municipalities in Iceland, there are normally no political factions in the sense that politics is based on persons and not on political parties. Hence, some of the smaller communities experience the work of political parties at the local level for the first time, which is decidedly different from the work of the previous arrangement found in the municipality.
Again, the main disadvantages are the reflection of the responses in the questionnaire. They point out that home rule councils do not function the same way in all municipalities. Hence, there is a question of the "local" in relation to the largest part of

STJÓRNMÁL & STJÓRNSÝSLA
The theoretical and practical dilemma of creating a new organisational design for a municipality using sub-municipal units municipality, previously called Fljótsdalshérað. Here, the home rule council does not have the obvious community connection as in the other three smaller villages. In several cases, it was pointed out that the home rule councils were overloaded with mandatory tasks, which prevented them from having a real effect on the locality. In addition, the place of the home rule council within the larger administrative organisation was unclear. '[The home rule council] needs a clearer position, it has in some cases lengthened the decision-making process, not always clear who has the final say.' Another respondent pointed out that the size of smaller communities was still a problem: Disqualification because of job or family connections is still a problem, and there is a real danger that when in conflict the representative that is appointed by the municipal council does not support the locality but the municipal council.
There were those who believed the arrangement had considerable flaws.
There are many flaws. Real issues are not sent to the home rule council. There is no trust towards the home rule councils to handle real tasks. Local planning should not be a task for home rule councils, it only increases the complexity in decision making and it is already very complex. An example of issues that should be with the home rule are harbour issues. The political [parties] are not ready to give up power to these local elected members [of the home rule council].
It was also pointed out that it was better to have someone from the local community as a leader of the home rule council, as that was the most efficient way to organise work. The possibility of multiple mandates was also criticised: a member of the municipal council could also be an elected member of a home rule council. It was also pointed out that the professionalism of the smaller communities was not strong enough, and in some cases, the new procedure was hollowing out the central administration. Moreover, it was also mentioned that it was not always clear what roles the municipal chief executive deputy hired in all the home rule councils should play within the wider organisation.

One municipality -many communities
A new organisational chart needed to be drawn for Múlaþing. Figure 3 shows the connections between different sections of the political arm. The design is a functional matrix with a divisional structure. The municipal council, with 11 directly elected members, is the key decision-maker. The following are three political committees: executive committee with five members covering finances and administrative matters; family committee with seven members covering education, social services, culture and recreation; and environmental committee with seven members covering all environmental issues, planning and building permits. Each of the four home rule councils has two directly elected members, in addition to one member who is appointed from within the munici-STJÓRNMÁL & STJÓRNSÝSLA pal council. Elections for the home rule councils take place parallel to the local elections. However, the leader of the home rule council is appointed by the municipal council from among its members.

STJÓRNMÁL & STJÓRNSÝSLA
The theoretical and practical dilemma of creating a new organisational design for a municipality using sub-municipal units

Discussion
There is a strong rhetoric for the virtue of the local (Purcell 2006) in all the preparation material leading up to the amalgamation referendum in 2019. This is again reflected in both the survey responses and the open responses. The respondents were, in general, highly positive towards the experiment. In sum, there is a clear logic of identity behind both the original ideas for the new municipal design and the respondents' evaluations of this design (Dardanelli 2021). At the same time, they pointed out some of the problems of this design, such as the power struggle between the municipal council and the home rule councils, a problem identified in previous studies on sub-municipal entities, as well as the problem of multiple mandates (Hlepas 2018). When looking at the organisational chart of Múlaþing through the lens of Mintzberg (1983), the chart reveals the strategic apex and the middle line of the municipal organisation, while the operating core, technostructure and support staff are mostly invisible. Around half of Icelandic municipalities have less than 1,000 citizens, and most function on a simple structure. Even a large municipality (in the Icelandic context) such as Múlaþing is still very small and highly questionable if the size of the municipality can sustain a divisionalised form. Divisional configuration is used when organisations need to adapt to local circumstances or would like to reduce risk for the overall organisation (Burton et al. 2021). In this case, support for the divisional configuration comes from the willingness to give a certain leeway to the individual needs of local communities. The positive side of this approach has been pointed out in the survey responses, as it does give certain autonomy to each division (or local community, in this case). It is also possible or even desirable, from a democratic point of view, to give different incentives to different divisions and thus create different outcomes. However, an important flaw of the divisional structure is how low it is in functional specialisation. There is also often a lack of coordination, which may lead to goal divergence and serious internal conflicts (Huczynski & Buchanan 2017). The current organisational design for the municipality of Múlaþing tries to tackle that problem in two different ways: (1) by appointing one member of the home rule council from within the municipal council and (2) by hiring a deputy for the municipal chief executive in each of the localities. In both instances, the aim is to establish communication between different layers of the government. This design also indicates a line of authority, an issue pointed out in the survey responses, in which a lack of authority on behalf of the home rule councils was questioned. The municipal council seems to be in a much stronger position within the organisational design than the home rule councils-a common setup in such designs (Hlepas 2018). On the other hand, there are those who question the virtue of the local and point out the danger of NIMBYism (Jun & Musso 2013). Furthermore, the position of the deputy of the chief executive seems to be more problematic than the situation of the home rule council itself. The deputy to the home rule council is a direct subordinate to the chief executive while working on a daily basis for the home rule council. The work of the home rule council is driven by identity logic, while the administrative work is driven by functional logic based on efficiency and effectiveness. It may be difficult to align both STJÓRNMÁL & STJÓRNSÝSLA ideologies. The findings demonstrate that the efficiency vs. democracy dilemma is very real in this setting. Based on this small survey it seems that this design may to some extent minimize the negative effect of the amalgamation in relation to local democracy and representation of local issues in a larger setting. On the other hand, it also reveals administrative problems such as the home rule councils being overwhelmed with tasks, possible low level of professionalism in decision making at the home rule level and unclear mandate between both the home rule council and the mother council as well as between individual staff members at different levels. Dardanelli (2021) argued that a government system has a given set of policy competences. Based on this argument, administrative decentralisation within a single municipality is only likely to fragment this specific set of competences. In sum, it will not increase the capacity of the municipality; it will just make coordination more difficult.

Conclusions
First indication of the Múlaþing experiment shows that political decentralisation has, to some extent, been successful. The organisational chart shows a divisional structure in which the locally elected arm has been successful in creating a legitimate ground for its existence. The main dispute seems to lie in the extensiveness of the self-control of the local units and in the perceived influence of the municipal council over the home rule council. One of the main objectives of amalgamation is to create a simple and effective administration. The findings demonstrate that this may be problematic. Both the organisational chart and the survey findings support the idea that the administrative capacity of home rule councils is weak. It is also possible that the administrative size of the municipality simply does not support a divisional structure. Consequently, the organisational design seems to be partly driven by the need to compensate for the loss of autonomy of previously autonomous communities. The rhetoric and slogans used in the preparation for the referendum support that conclusion, as do the survey respondents' emphasis on local self-decision-making.
This study is only the first step in the evaluation of the success of this project. More studies will have to be conducted along the way e.g., citizen survey as well as a systematic and longitudinal evaluation of the administrative consequences of the organisational design.
However, it is important for the municipality to take a hard look at its organisational design; the experiment is only three years into its eight-year process. There is still time to make changes and modifications that may improve the organisational functions of the municipality. This is especially important in relation to the fragmentation of municipal capacity when it comes to informed decision making and professional support to home rule councils.

STJÓRNMÁL & STJÓRNSÝSLA
The theoretical and practical dilemma of creating a new organisational design for a municipality using sub-municipal units