Contrary to prevailing trends in democracies of Western and Central Europe that have been consolidating their local government systems for decades, ex-socialist countries opted in many cases to fragment their local government territorial structure after the break-up of the old regime. Even among fellow countries in Central and Eastern Europe, Croatia stands out for the sheer scope of its local government territorial transformation during the period of transition. In only a few years after gaining independence in 1991, the number of municipalities more than quintupled. In relative terms, this constitutes the most radical case of local government fragmentation during a post-communist transition (Swianiewicz 2010; Baldersheim & Rose 2010; Swianiewicz & Gendźwiłł 2017). The focus of the paper is on the splits that happened after the initial “big-bang” of 1993 that was part of a complete redefinition of the local government concept and that was meant to be a clear sign of political departure from the previous system. The separations that followed the initial transformation are more easily observable and might reveal more concrete motivations and drivers that fuelled them. The fragmentation process in Croatia has not yet been thoroughly mapped nor subjected to an in-depth analysis and interpretation.
The goal of this paper is twofold: it seeks to analyse and explain both the process of fragmentation in Croatian local government in general and its variation among municipalities – individual cases of municipal splits. Therefore, analysis is structured along two intertwined research questions: 1: Which institutional and other contextual factors stimulated and facilitated fragmentation of the local government system? 2. Which local factors triggered individual secession initiatives? To understand these issues, the paper employs the explanatory framework developed by Erlingsson (2005) and adapted by Swianiewicz (2021), amending it with a couple of other contextual variables. The complementary influence of these factors contributed to a series of splits that happened between 1994 and 2006. To this day Croatia has been an exception among comparable countries by not having defined firm criteria and minimum population thresholds for the establishment of new municipalities (see: Swianiewicz 2021, Table 3). Therefore, all split initiatives have been heavily subjected to political arbitration and were materialised only when a balance between interests of local and national political actors was achieved. Understanding the explicit and underlying motivations of the actors involved, as well as the various institutional conditions that made the context favourable for secessions, might be useful in designing some future consolidation reform that would seek to avoid and prevent the negative effects of the large and centralised municipalities that provoked the aforementioned trends.
Contrary to amalgamation reforms that are often implemented forcefully and contrary to the will of the merging municipalities, fragmentation of territorial structure is predominantly bottom-up initiated. Still, without the favourable attitude of national political actors and a facilitating institutional and legal context, secession initiatives can hardly be successful. Secession initiatives emerge as a pursuit of the expected benefits of a fragmented local government. Such a territorial organisation has traditionally been advocated through several normative arguments, mostly pertaining to the issues of local democracy, but also to the allocative efficiency in delivering public services. Small communities have a tighter link between citizens and their representatives, which can result in easier control, and greater mutual trust and legitimacy. Small jurisdictions might be more accountable and responsive to the needs and preferences of the citizens and more easily provide opportunities for citizens’ participation in public affairs. Additionally, small-sized municipalities may provide better access to local services and allow for their tailor-made delivery (Baldersheim & Rose 2010; Swianiewicz 2010). However, the pursuit of these expected benefits is not universal, as secession initiatives emerge only in some municipalities and not in others. The local context that translates these arguments into impetus for secessions is therefore vital for understanding the emergence of secession initiatives. National governments might support municipal secessions for the same normative arguments, but their benevolent attitude towards fragmentation may also be shaped by their own gains, irrespective of potential negative effects on local government.
Arguably, all these elements are incorporated into Swianiewicz's proposed framework (2021). Following and adapting Erlingsson's model, which was developed for the understanding of municipal secessions in Sweden (2005), Swianiewicz integrates various variables into a single explanatory framework and employs both the analysis of institutional context of local government fragmentation and the examination of local conditions responsible for individual secession initiatives. The analysis of local drivers is based on three groups of economic, socio-cultural and political factors whose interplay may explain the pursuit of municipal independence. These factors almost always imply a certain heterogeneity within a municipality that produces centrifugal tendencies. Regarding economic factors, secession initiatives are expected to be more likely to emerge if there is a territorially defined part of a municipality that is richer, but not willing to subsidise poorer areas. The second line of economic reasoning presents secessions as a response to a feeling of being marginalised or exploited that can be developed in some parts of a municipality. Although the existence of this feeling and its relative importance for split decisions might be more precisely grasped only by a survey or a case study, it may be assumed to be real when indicated by several objective circumstances that increase the likelihood of it developing: great disparities in size between the “different” part and the rest of the municipality, peripheral location of a settlement and distance from the municipal centre or a situation of “no natural centre” in which no settlement is clearly dominant in terms of size. Cultural or identity issues constitute the second group of local drivers that implies socio-cultural differences within a municipality that can lead to its split. These differences usually arise in municipalities that incorporated culturally, linguistically or ethnically diverse areas; settlements that were historically administratively independent or those that are not connected by natural gravitational ties. The third group of factors relates to the political push for secessions. Political advocacy for secessions initiatives may be a mere articulation of the previously discussed factors by influential local political actors (officials of the existing municipality, politicians from the seceding part, national politicians interested in local issues for whatever reasons, citizens themselves), but it can also originate from a clear difference in political preference between different parts of a municipality, resulting in a quest for municipal independence.
During the first two decades of Yugoslavia (1945–63), local government territorial organisation was very unstable. Being vital for the realisation of “self-management” and “communal” concepts, local government was gradually subjected to a series of reforms that were usually quite drastic and often experimental in nature. The communal idea promoted large, self-sufficient municipalities, which were seen as the basis and primary entities for the political organisation of a society. Accordingly, the dynamics of local government territorial structure were unequivocally directed towards its consolidation. The number of basic local units in Croatia decreased constantly and substantially from 4.640 in 1945, through 737 in 1952 and 299 in 1955, to 111 in 1963. In most cases amalgamations were forced in a top-down manner, fuelled by political considerations and without (formal) consultations with citizens. However, they were not spatially arbitrary: in most cases, amalgamated municipalities did gather interdependent gravitational areas of urban centres rather precisely so their borders successfully embedded existing functional areas into single administrative jurisdictions. At the end of the consolidation process in the mid-1960s, Croatia had the largest municipalities in Europe at the time, with an average size of more than 40,000 inhabitants. In the following period, the local government structure stabilised and the number of municipalities changed only sporadically due to individual cases of territorial adjustments, so on the eve of independence and the change of regime, Croatia was divided in 102 municipalities (Žuljić 2001; Pavić 2010; Koprić 2010).
A couple of years after the change of regime and independence from Yugoslavia in 1991, Croatia radically departed from both the concept and territorial organisation of local government inherited from the communist period. At the beginning of 1993, a two-tier system was established, with 487 municipalities (417 rural municipalities –
In the years following the big break of 1993, the territorial organisation of Croatian local government continued to reshuffle, mostly incrementally. All cases of municipal secessions in the post-1993 period took place between 1994 and 2006, with one major wave of splits in 1997, two minor waves in 1995 and 2006 and several individual splits in-between (Table 1). All cases of splits were tracked by the official changes of relevant legal framework Although formally they represent different types of local units, towns and rural municipalities in Croatia have identical competences, with the exception of 25 bigger towns. Municipalities’ motives to “upgrade” their status were purely a matter of prestige and visibility.
Timeline and number of municipal secessions in Croatia
Number of splits | - | 1 | 1 | 49 | 4 | 2 | 1 | 1 | 1 | 2 | 6 |
Number of municipalities established | - | 1 | 10 | 52 | 4 | 2 | 1 | 1 | 1 | 2 | 6 |
Total number of municipalities | 487 | 488 | 498 | 539 | 543 | 545 | 546 | 547 | 548 | 550 | 556 |
In four cases, municipal splits resulted in more than one additional municipality, the most interesting being the case of Zagreb in 1995, when ten surrounding municipalities seceded from the city urban core. Unlike other splits, this case was only one in a series of administrative reshufflings that were performed on Zagreb and its surroundings in the period before and after independence. Municipalities that seceded from Zagreb in 1995 had been merged with the city as late as 1990. Since the merger had been done without any consultation with local institution or population and without historical precedent, these splits came as no surprise as they more or less re-affirmed the administrative boundaries that Zagreb had had before 1990 (Hrženjak 2000; Ivanišević 2000).
Not all newly established municipalities emerged from a single existing municipality – there were four cases of settlements being extracted from two existing municipalities. Six municipalities were split on more than one occasion in different years. All newly established municipalities, except two bigger towns that separated from Zagreb (Velika Gorica and Zaprešić), had significantly smaller populations than the current national mean. Furthermore, the vast majority of them (87%) were even smaller than the national median. Fourteen of them (17%) had a population below 1,000, which is a significantly bigger share of such municipalities than in the total population (6%). The population categories of new municipalities are presented in Table 2.
Population categories of new municipalities
Number of municipalities | 14 | 31 | 25 | 6 | 4 |
Share of new municipalities | 17.5% | 38.8% | 31.2% | 7.5% | 5% |
Municipal splits in the post-1993 period were not a mere continuation of previous systemic change, although they were certainly inspired by it. The initial positive effects of the previous fragmentation had outnumbered the negative – new municipalities managed to improve their basic infrastructure, and subsequent split initiatives certainly looked up to these positive examples. Also, after the big reform of 1993 that was implemented arbitrarily and without any (firm) criteria for the establishment of new municipalities, central government lost the moral grounds to prevent and deny further splits. The majority of splits in 1997 and later occurred in new municipalities – those that had been established in 1993 (51 cases) – and only a small portion of them implied (belated) secession from the old municipal centre (17 cases). Therefore, the reasons that led to the initial “big-bang” that were analysed previously are not automatically applicable to these later splits, nor sufficient to explain them. There are three groups of explanatory factors that shaped the institutional context favourable to stimulating secession initiatives and facilitative of their success: a) the legal framework for the establishment of new municipalities (criteria and procedure); b) the overall institutional position of local government within the politico-administrative system; and c) the political climate and considerations of influential political actors. The following paragraphs analyse each of these groups.
1. The territorial organisation of local government in Croatia has been a matter of an organic law that defines the number of municipalities and territorial reach of each. Therefore, every intervention in local government territorial structure requires a legislative procedure in the national parliament. General legislation on local government ( Although their consent is not required, these institutions could oppose such proposals most effectively by ignoring them, thus freezing any further procedure. Because of this, some of the split proposals were instigated even without these formal opinions (Hrvatski Sabor, 2005). Initiators were obliged to enclose a projection of expenditures of future municipality and adequate sources of revenue to cover them. This is problematic from the constitutional point of view, but it has not been officially addressed by the Constitutional Court. Such provision limits the constitutional role of parliament as the sole legislator who should be able to act independently from the cabinet as the executive institution.
2. The second wave of secessions in 1997 and following cases were heavily facilitated by the overall position of local government in the politico-administrative system. Such a position is predominantly indicated by the distribution of public functions among territorial levels (the functional aspect) and the fiscal regime of the local government (the financial aspect). The main purpose of local government was still a political one: to bring democratic institutions closer to the people and to engage them in the political process, while administrative and service-providing roles were secondary. Prior to 1993, arguably all local public affairs were performed by municipalities. From that year on, these tasks were distributed among newly established municipalities, counties, and a dense network of branch offices of state administration and other central public institutions. During the 1990s, Croatia was a heavily centralised and hierarchically governed country. In 1999, local government expenditure comprised only 6.5% of total public spending. Being relieved of many administrative and service-providing responsibilities, municipalities neglected capacity and efficiency considerations. In such a context, villages and small towns could “afford” secession and be on their own. The question of sufficient resources was rarely raised, as municipalities were not expected to do much – complex and expensive public services (healthcare, education, social welfare, fire protection) were the responsibility either of counties or of the central state. Even exclusively municipal services were not much of a concern. Those of economic nature (water services, collection of municipal waste, public transport) were in most cases delivered by companies that are formally inter-municipal, but effectively under the control of a bigger urban municipality in the area – usually an ex-municipal centre (Škarica & Giljević 2016). From 2000 onwards, the functional scope of local government was expanded, as decentralisation policies were implemented in the fields of education, social care, healthcare and fire protection. Although mostly limited to the counties and urban municipalities, this decentralisation coincided with the decrease in secession initiatives that has been seen since the beginning of the century.
Besides functional aspects, the financial context of local government also facilitated pursuits of municipal independence. In the second half of the 1990s, unconditional transfers (state aid) from national to local budgets comprised 8% of total local government revenue, this share being much bigger for rural municipalities (20%). Existing schemes of financial assistance to smaller and rural municipalities (the vast majority of seceded municipalities) provided a safety cushion to splitting settlements, thus reducing risks of a split – in a situation of financial deprivation they could count on additional resources. At the same time, personal income tax was the main source of local governments’ own revenue, belonging to the municipality of residence of the employee. Thereby, a seceding municipality could count on preserving its whole fiscal capacity no matter where jobs and businesses were located.
Another feature of the local government institutional context provoked municipal splits – the weak position of sub-municipal government institutions. In this period, sub-municipal government was completely marginalised and deprived of legal subjectivity, competences and finances. Such an institution could not serve its basic purpose – to articulate and to promote the interests and needs of individual settlements within municipalities. In this context, sometimes the only viable option for marginalised settlements was to establish their own municipality.
3. Democratic considerations and the predominantly (symbolic) political role of local government during this period were compatible with the political interests of major political actors. Despite justifying secessions by citing “the right of self-government” and “honouring the will of the citizens”, the attitude of ruling parties towards splits was primarily shaped by political gains: the more municipalities there were, the more political seats there were to win. At the same time, a fragmented local government structure could secure the continuation of a centralised governance model and help to preserve political power within the hands of national political elites. The vast majority of splits occurred during the terms of centre-right governments led by the conservative Croatian democratic union (
Municipal secessions can be either facilitated or hampered by the national institutional context, but they are primarily pushed by a complex interplay of various local factors. This section follows the proposed framework of Swianiewicz (2021) and summarises a multitude of local factors and their possible indicators in three distinctive groups – economic, cultural and political. It examines the influence of all the proposed indicators on which data were obtainable by desk research. The following analysis deals with 63 cases of municipal splits, which resulted in 66 newly established municipalities In three cases, two municipalities seceded at the same time from the abandoned one.
1. In general, most of the splits occurred in economically deprived areas. Only 13 pre-split municipalities (20%) had
Ratio of per capita expenditure between new and abandoned municipalities All financial indicators are drawn from the annual financial reports of local government units compiled and published by the Ministry of Finance.
Number of municipalities | 9 | 28 | 20 | 9 |
Share of municipalities | 13.6% | 42.5% | 30.3% | 13.6% |
Another line of economic argumentation follows the subjective feeling of a “different part” in a municipality being marginalised or exploited. As was previously discussed, this feeling may be developed under various circumstances of structural differences between the seceding and remaining part of a municipality. Regarding difference in population size between new and abandoned municipalities, almost half of the Croatian cases of splits belong to the “even-steven” category, according to Swianiewicz's proposed typology (2021), while remaining cases are almost equally distributed between “grown-up leaving the nest” and “nestling leaving the nest” groups (see Table 4). There were even nine cases in which a seceding municipality had a bigger population than the abandoned one, but not by much, so these cases are classified into the “even-steven” category. Mean population size ratio between the abandoned and new municipalities is 4.6.
Population size of new municipalities in relation to the abandoned
Number of municipalities | 31 | 18 | 17 |
Share of total splits | 47% | 27% | 26% |
As with previously discussed financial indicators, there is a clear dividing line between municipalities that seceded from the old centres and those that split from municipalities established in 1993. The mean population size ratio between municipalities in the first group is 10.1, while in the latter case it is only 2.3. Fifteen of 19 municipal splits within the first category resulted in the establishment of “nestlings” – municipalities smaller than one fifth the size of the abandoned one – which constitutes 88% of all such cases. Conversely, out of 47 secessions in the second group, 31 were “even-steven” cases (all such cases are from this group of municipalities), 12 are “grown-ups leaving the nest”, and only four represent cases of nestlings. Marginalisation of “nestlings” does not stem only from their smaller size, but also from the fact that these cases of splits occurred in spatially large municipalities in which a sense of marginalisation in peripheral parts was more easily developed. Pre-split municipalities that nestlings seceded from have an average area much larger (125 km2) than both the national average before 1997 splits (114 km2) and the current one (102 km2). A large number of even-steven cases shows that territorial reform in 1993 produced many municipalities with no natural centre. Settlements of similar size have found themselves in centre–periphery and superior– subordinate relations never experienced before, albeit all of them belonged to the same large pre-1993 municipalities. Naturally, these situations produced centrifugal tendencies as none of them had sufficient gravitational force or economic strength to preserve the unity of their municipality. These cases were especially frequent in economically lagging parts of the country where seceding parts sought an opportunity to improve their conditions by establishing direct connection to national fiscal equalisation schemes by having their own municipality.
Another indication of the importance of the marginalisation argument for the analysed secessions is the surprisingly big share of splits in which only one settlement separated. Croatian municipalities currently comprise approximately 12 settlements on average and there are only 45 (8%) municipalities with only one settlement. Even 26 of them resulted from these splits (58%), which means that in 40% of the analysed cases, a single settlement separated from the remaining part of the municipality. The number of settlements in municipalities abandoned by the single settlement varies from one (in two cases) to 32. In a way, separations of single settlements not only indicate the economic reasoning of marginalised areas, but also show their distinctive identity.
2. Among cultural factors and identity issues, reminiscence of having been their own separate municipality might have some effect, but its importance is difficult to discern. Apart from the last three decades of the communist regime, Croatia traditionally had a rather fragmented local government structure. Arguably all separated municipalities might have invoked a period in which they were an autonomous entity. However, there are at least two circumstances that diminish the relative weight of this factor: a) the most recent cases of separate autonomy preceded these splits for 30–40 years and date back to the pre-1963 period, which brings into question the importance of such a distant memory for modern-day decisions; b) in 1945–55, arguably all current municipalities (including those that seceded) were separate entities, as local government was founded on the extremely large number of basic units, which in 1952 was rationalised to 737. If a memory of that period played a significant part in explaining these splits, there would be a much bigger number of secession initiatives.
The issue of wrongly drawn borders having connected different identities does not play a major role in Croatian splits. All municipalities involved in splits (new and abandoned) were within the boundaries of the same socialist municipality for at least three decades, so none of them were put together for the first time by the local government reorganisation in 1993. Cultural, linguistic or ethnic/religious differences among municipalities that divorced were and still are almost non-existent. In only one case is there a comprehensive difference in ethnic composition between the new and the abandoned municipality. This is the case of Negoslavci municipality, which has an overwhelming Serb majority that seceded from the municipality of Bogdanovci in 1998.
3. The influence of local political factors on municipal splits in Croatia seems to be of moderate importance, at least according to the indicators that could be obtained. Municipal secessions might be pushed by territorially defined differences in political preference within a municipality. As secession initiatives never have a formal partisan affiliation and there are no voting records for different parts of municipalities on elections that preceded the split, the weight of this factor can be examined only retrospectively. If territorially defined political opposition was decisive for a split, one would expect this circumstance to materialise as different outcomes of local elections following the split in new and abandoned municipalities. However, this difference was manifested in only a minor share of electoral outcomes. In 51 cases (77%), the first local elections after the split were won by the same party in both new and abandoned municipalities, and only in 15 cases (23%) did the electoral outcome produce different winners. The results of subsequent local elections (the second cycle after the split) show a more divergent picture – in 27 cases (41%) there is a difference between the new and abandoned municipality regarding the most successful party at local elections. For the purpose of this argument, the winning party in a municipality is indicated by the relative majority of votes in local council elections. Electoral results are obtained from the archive of local elections’ official results at the national electoral commission website:
The dynamics of the Croatian case of local government fragmentation do not deviate much from other cases in CEE countries. A favourable political and legislative environment during the early transition period produced a multitude of realised secessions. A common urge was to distance the new state and its institutions from the old regime as radically as possible. In the local government context this meant that fragmentation was not only tolerated but desirable. At the same time, secession initiatives were dealt with in a politically arbitrary manner, without any general guidance or criteria. The analysis has shown the gradual change in both political attitude and formal rules that slowly but steadily closed the window of opportunity for further splits. The absence of new splits coincided with the tightening up of procedural rules. However, the reasons for this are of a more political nature – every government since the beginning of the new century has been under severe public pressure to rationalise the politico-administrative apparatus. A combination of international (EU) pressure and recommendations, critique from the academic community and economic recession decisively contributed to the absence of splits. There have been at least a dozen initiatives in recent years, but none even reached the parliamentary arena.
Despite the arbitrariness and voluntarism behind their emergence, the analysed cases of splits have shown a certain level of common reasoning behind them. To a certain extent all examined types of factors contributed to individual cases of municipal secessions and are helpful in understanding their manifestations. However, the analysed data have given the biggest support to economic considerations having been behind the splits. Economic deprivation in general provided a big push for secessions. Two types of municipalities were susceptible to splits mostly – large municipalities with smaller and more distinctive peripheral parts, and small-to-intermediate rural municipalities with only a few settlements of similar size and importance.
The problem of territorial fragmentation and its consequences have been noticed and discussed extensively over the past two decades. The period of recession and stagnation (2009–14) brought an opportunity not only to debate, but to commence a profound structural reform of the local government system (Koprić, Škarica & Manojlović 2019). Still, no recent governments have wanted to engage in a process of territorial consolidation. On the other hand, they were equally reluctant to allow further fragmentation. It seems that there is a tacit consensus that the number of municipalities is not to be increased, and preserving the