The Scope of City Autonomy in the Constitutions of the Netherlands and the United Kingdom: Informality, Subsidiarity, Identity

One of the main issues in the debate on urban constitutionalism is how constitutions can recognize the increasingly important role of cities in relation to the nation-state. This paper examines what we talk about when we talk about city autonomy. This is a pressing question, particularly in the context of European unitary states. This paper pays special attention to the context of two of such states, namely the Netherlands and the United Kingdom (especially England). First, it explores the notion of subsidiarity, which implies that consideration should be given to the distinctiveness of the city as regards the allocation of power to the central and regional levels respectively. However, this idea in itself cannot justify the case for city autonomy, as the claim that the attribution of autonomous powers to cities may improve the quality of decision-making in the state as a whole needs additional empirical evidence. Second, it investigates the concept of city autonomy by exploring the fuzziness of the notion of the city. In addition, it introduces the concept of ‘spatial identities’ in order to explain the interdependence of (large) cities and their surrounding (rural) areas. Lastly, it concludes that if the importance of cities as constitutional actors is to be increased, it should be done so in an informal way rather than by the introduction of formal constitutional arrangements both from a pragmatic and a normative perspective.


Introduction
Urbanization provides an interesting puzzle for constitutional scholars. The world population increasingly resides in a city. In 2019, already 56% of the world's population could be described as city dwellers. 1 This proportion is due to rise to 70% by 2050, which makes policymaking even more dependent on the well-functioning of urban agglomerations worldwide. 2 In comparison to nation-states, cities generally share a relatively manageable scale and a relatively dense and diverse population, which arguably makes them able to tackle substantial policy issues better than nation-states. This could make the case for a power shift from the traditional nation-state to (large) cities thereby recognizing the latter as autonomous constitutional actors. 3 If this analysis were adopted, then the very essence of most constitutions appears to be very problematic. Hirschl, for instance, notes that most cities are dependent on 'constitutional structures, doctrines, perceptions, and expectations that were conceived along with the modern nation-state'. 4 On his account, the 'urban age' urges to 'remedy' the 'constitutional non-existence' of cities not so much 'via legal byroads, but rather through public law's main highway: constitutional law'. 5 Apart from the potential value of Hirschl's idea in an abstract sense, 6 it is not easy from a more practical perspective to imagine what a constitutional arrangement regarding the position of the city should look like. One of the first challenges is to define the very notion of 'cities', which is inherently unsettled. First, the exact boundaries of cities are in general unclear: they can correspond with the city limit (the city proper), they may be a congregation of multiple towns, cities and suburbs with a high population-density (urban area) or they may also include its surrounding territories and thereby comprise multiple (larger and smaller) cities, jurisdictions and municipalities (the metropolitan area). In the remainder of this paper, for the sake of convenience I will use the term 'city' to describe regions that are both (highly) urbanized and operate as a recognizable unity.
After surpassing the hurdle of defining the city, another complicating matter arises. This is, if cities are to be constitutionally recognized, how justice can be done to the very different urban contexts that exist around the world. Cities come in all shapes and sizes: the dynamics of rapidly expanding megacities with more than 10 million inhabitants, such as Bangkok and Beijing, 7 are truly different from those in Oslo and Amsterdam, being the largest cities in Norway and the Netherlands respectively despite having less than 1 million inhabitants. 8 It seems unlikely that, apart from London, megacities of more than 10 million inhabitants will emerge in Northern European countries, such as the United Kingdom, the Netherlands and 2 Ibid. 3 See, e.g., Barber (2013), p. 5; Hirschl (2020), p. 222; Adams et al. (2017) at 2728-2729. 4 Hirschl, City, State, p. 9 (n 3). 5 Hirschl, City, State, p. 16 (n 3). 6 Hirschl's ideas about the importance of cities to global governance will not be discussed in this paper. See for a critique on Hirschl's account of 'the sovereign city' as being 'subjugated by a Westphalian sovereigntist order' (see p. 10 of his work in the previous footnote) Loughlin (2022) at 357-361. 7 In the Bangkok metropolitan area reside approximately 16.255.900 inhabitants (https:// www. cityp opula tion. de/ en/ thail and/ cities/ ua/, 1 January 2022), while the metropolitan area of Beijing has a population of 21.893.095 inhabitants (https:// www. cityp opula tion. de/ en/ china/ cities/, 1 January 2022). 8 The municipality of Amsterdam has a population of 921.468 inhabitants (https:// www. cityp opula tion. de/ en/ nethe rlands/ admin/ noord_ holla nd/ 0363__ amste rdam/, 1 January 2023), while 699.827 people reside within the borders of the municipality of Oslo (https:// www. cityp opula tion. de/ en/ norway/ admin/ oslo/ 0301__ oslo/, 1 January 2022).
the Scandinavian countries. The relatively equal distribution of wealth across this part of Europe and the relatively fuzzy urban/rural divide implies that much, if not all, land use 'is already determined by urbanizing demands '. 9 Thus, it is very much dependent on the specific constitutional context of a country whether a specific area could be seen as (part of) a city that needs to be constitutionally recognized or not. That is not to say that the constitutional non-existence of cities has to be rejected all the way. Indeed, even these highly urbanized states face challenges which could possibly be best solved on the city level. This paper therefore aims to explore whether the idea of the constitutional recognition of 'cities' could be modified in such a way that it would fit the specific context of the Netherlands and the United Kingdom. Both countries together help to gather broadly applicable insights for other (European) countries with similar characteristics.
The argument offered here has three steps. First, it will be explained that it is far from self-evident that the constitutions of both countries would entrench the autonomous position of (particular) cities. Both countries are unitary states which are largely urbanized, 10 but do not as such attribute significantly more powers to their cities (except for Greater London) than to other (rural) regions. Therefore, if the special position of (large) cities were to be recognized constitutionally, informality is key. This also suits well with the character of the institutional arrangements in both constitutions as being the result of organic developments rather than of a deliberate design.
Second, the rule of law implications of urban constitutionalism in both countries are explored. For the purposes of this paper Tamanaha's model of the formal version of the rule of law will be followed. If this version is considered in the context of the Dutch and the UK constitution, then its thickest formulation can be applied. Indeed, from the perspective of the rule of law, a constitutional arrangement involving city empowerment should at the very minimum be founded on the consent of the citizens involved ('democracy + legality'). 11 It is uncontroversial that both requirements are inherent to the conception of the rule of law as adopted in the constitutions of both the Netherlands and the United Kingdom. 12 It will appear that a formal power shift from the national level to the city level does not immediately fit well with the notions of democracy and consent as inherent to the rule of law. 9 Loughlin, 'The City in the Constitutional Imagination' (n 6). See also Arban (2021) at 346. 10 92,57% of the Dutch population and 84,51% of the inhabitants of the UK lived in an urban area in 2021, according to the statistics provided by the World Bank. See https:// data. world bank. org/ indic ator/ SP. URB. TOTL. IN. ZS? most_ recent_ value_ desc= true (last accessed on 11 March 2023). 11 These are the second and the third version of the thin conception of the rule of law in the model presented by Tamanaha. (The first version of this conception, which involves that the very form of law is used as a basis for government action, is inherent to the other versions of the thin conception of the rule of law.) See Tamanaha (2004), pp. 93-100. I will not discuss the implications of urban constitutionalism for the thick conception of the rule of law in this paper, as the applicability of this conception is more controversial. 12 See, e.g., the discussion of the legality principle and the notion of democracy in Kortmann et al. (2021), pp. 363-372 (for the Netherlands); Elliott and Thomas (2014), p. 64;Bradley et al. (2015), pp. 81-86 (for the UK).
The justification for the constitutional empowerment of the city is usually found in the notion of subsidiarity. Subsidiarity may appear to be a form of doing justice to the distinct position of the city in the nation state (the level of the city is most suitable to tackle particular policy issues in comparison to other government levels). This argument could only warrant the constitutional recognition of cities if it meets the standards of democracy and consent. More specifically, the idea of constitutionally empowering cities requires the establishment of a democratically vibrant community on the city level. 13 This entails that a power shift from the central to the city level should rest on the consent of all citizens affected, so both those from within and outside the borders of the city region involved. Therefore, the allocation of specific powers to the city level should respect the interconnectedness between urban and rural interests.
Third, the concept of spatial or regional identities will be introduced as a model. This concept, drawn from the work of the political geographers Anssi Paasi and Kees Terlouw, is a social construct which exists of so-called thick (historically and institutionally embedded) and thin (future oriented) identities. It involves a theoretical notion that may help to understand how urban and rural interests within a nationstate can be aligned as it stresses the interdependence between larger (metropolitan) cities and their surrounding (rural) areas.
The structure of this paper is as follows. In Sect. 2, the basic structure of the constitutions of the UK and the Netherlands will be briefly explored. Section 3 explores the relation between the subsidiarity principle and the position of cities in both constitutions. In Sect. 4, the concept of spatial identities will be introduced and applied to the position of cities in both the Netherlands and the UK. Section 5 contains some concluding remarks.

The Structure of the Constitutions of the UK and the Netherlands: Informality as Guiding Principle
This section explores the structure of the constitutions of the UK and the Netherlands, which helps to appreciate the position of cities in both countries. Both constitutions leave considerable leeway to political rather than judicial institutions as to how constitutional values should be protected. This is the result of their lack of a 'grand design'. In this section, I offer a short overview of the most important characteristics as regards the structure of the UK constitution (Sect. 2.1) and the Dutch constitution (Sect. 2.2). The remainder of this section contains some remarks on the nature of politically enforced constitutions (Sect. 2.3).

The UK Constitution
The famous constitutional scholar A.V. Dicey (1835Dicey ( -1922 has famously characterized the uncodified UK constitution as 'historical'. 14 By using this expression, Dicey meant that the design of the UK constitution has never been the result of a deliberate choice, simply because there has never occurred a revolution or war in this country that has forced the constitutional law-maker (whatever its composition would have been) to enact a codified Constitution. The cornerstone of the uncodified UK constitution still remains the sovereignty of Parliament. This doctrine, following its orthodox conception by A.V. Dicey, holds that 'Parliament has the right to make or unmake any law whatever and, further, that no person or body is allowed to override or set aside statute law which Parliament has adopted'. 15 Therefore, the UK constitution is usually regarded as extremely 'flexible', as Parliament is unlimitedly competent to change any constitutional arrangement. It can simply do this by enacting an ordinary Act of Parliament, without having to follow a specific constitutional revision procedure. 16 Although the Diceyan conception of this doctrine has become increasingly hard to sustain due to the introduction of the devolution settlements and the enactment of the Human Rights Act, 17 it can still be regarded as the source of a constitutional culture in which there is a strong judicial reluctance to interfere in the political or legislative process. 18 One fundamental qualification should however be made here: Parliament cannot change the terms of its own power, for instance by abolishing or significantly changing the very doctrine of the sovereignty of Parliament. 19 Indeed, it is not even clear whether this doctrine can be terminated at all. In this sense, the UK constitution is, when it comes to the distribution of powers between the state institutions, extremely rigid. 20 Thus, the introduction of new settlements in the UK constitution is only possible insofar it does not formally interfere with the sovereign position of Parliament. The attribution of specific powers to cities remains conditional upon Parliament's willingness to respect this new distribution of power. If a new institution on the city level were established, then Parliament, due to its sovereign position, would remain perpetually authorized to override or to alter its powers or even to abolish it altogether. Apart from the procedure for the enactment of Acts of Parliament, the main rules that govern the relations within and between political institutions can be considered as constitutional conventions. They impose extra-legal constraints on the power of (constitutional actors within) Parliament to which it considers itself bound. 21 The difference between formal (legal) and informal (extra-legal) constitutional 14 Dicey (2013) at 172. 15 Dicey (1959), pp. 3-4. 16 See, e.g., Bradley, Ewing & Knight, Constitutional and Administrative Law (n 12), p. 7. 17 See, e.g., Barber (2011), pp. 144-154;Bogdanor (2019), pp. 51-86. 18 See, e.g., Geertjes and Uzman (2018), pp. 89-90. 19 This problem was notably exposed in R (Jackson) v Attorney General [2005]  arrangements should however not be overstated. In the UK context, the term 'law' refers to those rules that are enforced by the courts. 22 The non-legal character of conventions does not so much reveal anything about their bindingness. It does however underline that conventions are political rules, which cannot be enforced by the courts. In the political process, it should be determined whether political institutions have followed these conventions properly.
For the purposes of this paper, the most important of these political understandings is the devolution settlement. In this settlement, the UK Parliament established parliaments and governments with their own autonomous powers for the non-English parts of the UK (Scotland, Wales and Northern Ireland). Since 1998, new relations between the UK Parliament and the Scottish Parliament have been governed by the so-called Sewel Convention, according to which 'Westminster would not normally legislate with regard to devolved matters in Scotland without the consent of the Scottish Parliament'. 23 One year later, in a Memorandum of Understanding between the Westminster Parliament and the devolved institutions in Wales and Northern Ireland, a similar convention was announced as regards the relationship between the UK Parliament and the other two non-English parts of the UK: the Legislative Consent Convention. The introduction of some form of city autonomy (to some extent, the introduction of the Greater London Authority could be seen as such) 24 requires the introduction of such a devolution settlement.
Although the devolution settlement has been designed in such a way that the sovereignty of Parliament remains legally intact (the UK Parliament holds the legal right to enact legislation for the devolved parts of the country), from a political perspective it has become increasingly difficult for this institution to ignore the rights of these devolved institutions. 25 The partial codification of the Legislative Consent Convention has underlined (albeit without granting an explicit legal status to the Convention) that the powers of the UK Parliament have significantly shifted to the devolved institutions in Scotland and Wales respectively. 26 The central position that the UK constitution still attributes to the sovereignty of Parliament has left the UK's character as a unitary state untouched. In reality, however, the underlying conventions of the devolution settlement have, as Barber notes, appeared to be effective and it seems unlikely that the UK Parliament will abolish the devolved institutions without their consent. 27 In this realistic sense, the UK is increasingly resembling a federal state in which the powers of both the central government and the devolved governments are constitutionally recognized.

The Dutch Constitution
The constitution of the Netherlands can be similarly described as a document that was 'not made but grew'. Although the Netherlands has the second-oldest codified Constitution in the world, 28 a strong constitutional tradition has never come to rise in this country. 29 This is due to 'the very austere quality' of its provisions: the core of most of these provisions dates back from 1848 or 1917, having not been fundamentally revised after these years. Moreover, the political nature of the Dutch Constitution is famously expressed by Article 120 (dating back from 1848), which holds that the constitutionality of Dutch Acts of Parliament (wetten in formele zin) shall not be reviewed by the courts. Although the courts have been authorized for almost seven decades now to review the compatibility of Acts of Parliament with international human rights treaties including the ECHR, 30 Article 120 can still be seen as a proof that matters of constitutionality (including institutional issues) are ultimately of a political nature, best left for political institutions.
No Dutch institution is above the Constitution. If any institution has to be designated as sovereign, then this should be the constitutional lawmaker. In the context of local government for instance, the Constitution entrenches the position of the most important local authorities, namely the municipalities (gemeenten), provinces (provincies), and local water authorities (waterschappen). 31 These provisions attribute two types of powers to these local government institutions, namely, 'delegated' powers (bevoegdheden uit medebewind) which local authorities have to perform by order of the central government 32 and 'autonomous' powers (autonome bevoegdheden) which can be found in the Province Act (Provinciewet) and the Municipality Act (Gemeentewet), which local authorities can wield on their own accord. 33 The importance of the distinction between these two powers is limited, notably as it is the parliamentary law-maker on the national level (the formele wetgever) 34 that determines what the autonomous powers of the local government institutions involve.
The Constitution principally recognizes the mere existence of local government institutions and acknowledges that they should at least have some autonomous powers regardless of the desires of the central government. This is important, as 28 Usually, the Dutch constitutions is regarded as the third-oldest constitution after the constitutions of the United States and Norway. However, the Norwegian constitution was enacted on 17 May 1814 at Eidsvoll. The Dutch constitution was established on 29 March 1814 and is thus 6 weeks older than the Norwegian one. See more extensively Voermans (2018) at 13. 29 Voermans, 'A 200-Year-Old Constitution: Relic or Enigma?', p. 12 (n 28). 30 See the Articles 93 and 94 of the Dutch Constitution. 31 See particularly the Articles 124 and 133 of the Dutch Constitution. The Dutch Constitution also recognizes three 'Caribbean public bodies', namely Bonaire, Sint Eustatius and Saba (see Article 132a of the Dutch Constitution), but that is not relevant for the purposes of this paper. 32 Article 124(2) of the Dutch Constitution. 33 Artikel 124(1) of the Dutch Constitution. 34 The formele wetgever is constituted by the government (the regering) and Parliament (the Staten-Generaal). See Article 89 of the Constitution. As opposed to the UK Parliament, members of the Dutch government do not have a seat in the Dutch parliament. Thus, the position of the government and Parliament altogether is somewhat comparable to that of the UK Parliament.
the Dutch Constitution is considered to be rigid. 35 Article 137 of the Constitution provides that any bill that proposes an amendment to the Constitution has to be accepted by both Houses of Parliament in two stages. In the first stage, such a bill has to be accepted by a majority of the members of both Houses of Parliament in order to pass. Subsequently, the dissolvement of the Lower House (the Tweede Kamer) and new elections (which in fact are both always combined with the regular dissolution of the Lower House) have to take place, after which the second stage may be initiated. In this stage, both Houses of Parliament need to vote again on the same bill that proposes the amendment of the Constitution. The proposed revision of the Constitution is only successful if a two-thirds majority of the members in both Houses of Parliament accept the bill. Although this procedure seems relatively straight-forward, it places very high impediments for proposals that intend to significantly revise the Constitution. Only relatively uncontroversial proposals for the amendment of the constitution are likely to pass these hurdles. 36 It would be hard to think of a proposal for constitutional revision aimed at the enhancement of city powers in relation to the central government that would not be seen as sufficiently uncontroversial to make it to the Constitution.

Politically Enforced Constitutions and the Limits to Constitutional Reform
The very limited overview above shows that the development of the Dutch and the UK constitution is highly dependent on factual and political developments rather than on a deliberate legal design. Such a piecemeal approach of constitutional development may be seen in positive terms in the sense that it generates little conflict: by only entrenching the most fundamental rights and the mere existence of the most important institutions in the Constitution (the Netherlands) or by simply not codifying the constitution at all (the UK) it is hard to disagree with the substance of a constitution as such. The downside of this structure is that it significantly limits the possibilities for constitutional amendment. If a factual development urges to fundamentally alter the constitutional framework, the central position that the constitution attributes to the central Parliament (in the UK) or the rigid character of the constitution (the Netherlands) appears to be obstructive.
This need not be problematic. Generally, it could be argued that a proposal for constitutional reform needs to be based on sufficiently broad and consistent consensus. This is important from the perspective of what Barber calls 'good-enough-constitutionalism'. Barber states that radical changes to the framework of the constitution such as the replacement of a monarchical for a presidential system may have a destabilizing effect on constitutions. 37 Despite the inherent merit of a proposal, it may have unintended consequences that cannot be completely foreseen. It could for instance appear to be difficult to agree on what the alternative to a monarchical system should look like. The problem of unintended consequences of the amendment 35 See, e.g., Gerards (2016) at 225. 36 Gerards, 'The Irrelevance of the Netherlands Constitution', pp. 225-226 (n 35). 37 Barber,The Principles of Constitutionalism,. of the constitution, albeit less radical than the abolishment of the monarchy, may similarly be applied to the idea of the constitutional entrenchment of city autonomy. In an abstract sense, the constitutional empowerment of cities may seem attractive. It is however not clear at all how this idea should be implemented and what its implementation costs will be. It should first be considered to what extent there is a need to constitutionally recognize the position of cities in the constitutions of both countries. If there appears to be such a need, then informal ways to do justice to the city's potential for the constitution should be favored.

Cities in the Constitutions of the UK and the Netherlands from a Rule of Law Perspective: Subsidiarity as a Lens
In this section, I will take a closer look to the possibilities for the recognition of the autonomous position of cities that the constitutions of the UK and the Netherlands could offer from a rule of law perspective. The thick and formal rule of law conceptions in both constitutions require that the content of the law should be determined by the consent of the citizens involved. 38 A potential constitutional arrangement relating to the relationship between the city and the central government level thus needs, from a rule of law perspective, a legal arrangement proving the consent of the citizens involved. This idea manifests itself more particularly in the idea of subsidiarity, which I will use in this section as a lens. First, I explain what subsidiarity generally entails. (Sect. 3.1). The remainder of this section contains an analysis of the need for constitutional recognition of cities from the perspective of subsidiarity for the constitutions of the UK (Sect. 3.2) and the Netherlands (Sect. 3.3). Due to the organic development of both constitutions, I specifically aim to determine to what extent the recognition of cities as autonomous constitutional actors is possible without the need of formal entrenchment.

The Notion of Subsidiarity
The case for city autonomy is usually informed by the notion of subsidiarity. 39 This notion has notably been shaped in the EU law context, where it is formally recognized as a legal principle. According to Article 5(3) of the Treaty of the European Union (TEU), the European Union shall act only in areas which fall outside its exclusive competence, 'if and in so far as the objectives of the proposed action cannot be sufficiently achieved by the Member States, either at central level or at regional and local level, but can rather, by reason of the scale or effects of the proposed action, be better achieved at Union level'. This provision is complemented by Article 1 of the TEU, which more generally articulates that government actions in the Union shall be taken 'as closely as possible to the citizen'. 38 Tamanaha In the context of city autonomy, two forms of subsidiarity can be generally distinguished: subsidiarity-as-respect-for-distinctiveness-of-the-city and subsidiarityas-democracy. 40 The first form of subsidiarity involves that the city's distinct position with its own distinct issues, due to its high density, its high diversity and its highly complex economy, warrants the direct attribution of specific powers to this level of government. As particular national challenges, such as migration, health care and distribution of food, have the most impact on the level of the city, this could generally make the case for the direct allocation from the national to the city level of powers to address these issues. 41 Although the notion of subsidiarity seems to emphasize the importance of city autonomy at first sight it cannot in itself provide a general justification for it. The notion of subsidiarity is 'normatively empty' because it neither addresses how the cities' boundaries should be drawn, nor how to ascertain which powers should be allocated to this level of government. 42 It is not a good in itself to protect authorities of the local government level, such as cities. The subsidiarity principle only involves the attribution of power to a subnational (city) level if the national authorities are not sufficiently able to achieve a particular objective of government action. If this is the case, the second category of subsidiarity, subsidiarity-as-democracy, comes in. Subsidiarity structures the democratic process by demarcating the boundaries between democratic units within a constitutional system. Therefore, it requires that newly created legislative bodies on the city level are democratically vibrant. The notion of democracy thus informs the subsidiarity principle. Barber contends that the creation of effective democratic units is dependent on the presence of social solidarity within that unit. 43 Democracy involves deliberation, and deliberation is only possible if all members of a specific democratically organized group are willing to listen to each other. The establishment of a new democratic unit demands that the members within that unit are sufficiently willing to value each other's interests.
Generally, this makes the case for maintaining the status quo rather than creating new governmental bodies on local levels. If a new democratic unit were established, the social solidarity that is already inherent to the pre-existing democratic (national or local) units may form a prima facie argument to leave the existing vertical distribution of power within a state untouched. The fact that boundaries between particular units already have been in place for a long time may gather a shared experience 40 Cahill and O'Sullivan, 'Subsidiarity and the City', pp. 58-68 (n 13). The authors separately distinguish between 'subsidiarity-as-efficiency' (the city level may be the most efficient level within the nation state to tackle specific issues) and 'subsidiarity-as-respect-for-distinctiveness' (the distinctiveness of cities, due to their dense complex integration and proximity, makes the case for the allocation of specific powers to the city level). As the authors implicitly acknowledge in their chapter (see p. 65), these forms overlap with each other. Therefore, I will combine them and discuss both forms simultaneously as if they are just one category of the subsidiarity argument. 41 Adams et al., 'Constitutionalisme in de eeuw van de stad', p. 2728 (n 1). 42 Barber,The Principles of Constitutionalism,. See also Muñiz-Fraticelli (2014), p. 70 citing Levy (2000) at 462. 43 Barber,The Principles of Constitutionalism,p. 195 (n 27), citing Song (2012) that could help a group of people to accept these very boundaries. That is not to say that the boundaries of particular units may never be changed. Social solidarity within a democratic unit may of course grow after its establishment. This is possible if there is an urgent need to change the existing borders that may make this social solidarity of secondary importance or if these borders may no longer have any significance at all. In that case, the generally distinct position of cities may inform the notion of subsidiarity-as-democracy. Cities are characterized by, firstly, a relative shortage of space and, secondly, by an economic dependence on big businesses, which may result in distinctive concerns for instance as regards the right to housing and inequality. For Hirschl, these characteristics illustrate the 'super-diversity' that he considers to be unique to the city. 44 Against this background, the question raises whether the distinctiveness of the position of cities within the constitutional contexts of the UK and the Netherlands warrants the constitutional recognition of their position. The subsidiarity principle requires clarity as to how the borders of cities should be drawn from a constitutional perspective. 45 At the very least it should be possible to determine whether autonomous powers should be attributed only to cities with more than ten million inhabitants ('megacities') 46 or also to other cities with fewer inhabitants.
The remainder of this section aims to answer the question to what extent it is possible to do justice to the distinctiveness of the position of cities within the constitutional contexts of the UK and the Netherlands. Importantly, my concern is not so much whether the specialness of cities in itself requires the attribution of powers to the city level from a democratic perspective. Rather, I aim to investigate for both countries if constitutional city empowerment could be established without distorting the balance between urban and rural interests within the whole nation.

The Potential for the Recognition of Cities in the UK Constitution
The current composition of the United Kingdom has been the result of a long historical process during which ultimately three nations, namely Wales (1536/1543), Scotland (1707), and Ireland (1800; in 1921 followed by the secession of all Irish counties which constitute the current Irish Republic, except for the six counties in the north east that constitute the province of Northern Ireland within the UK) ultimately joined England in an incorporating union. 47 As a result, the development of the constitution of the UK could to a large extent be seen as the continuation of the English constitution. The only part of the UK which does not have a place in the devolution framework of 1998, is England. The Westminster parliament is still dualhatted: it is the legislative body for both England and the overarching UK. According to Bogdanor, the main reason for England being left out of the devolution settlement, is that this precluded the non-English parts of the Union from seceding from 44 Hirschl, City, State, p. 226 (n 3). Cahill and O'Sullivan, 'Subsidiarity and the City', pp. 65 (n 13). 45 Cahill and O'Sullivan, 'Subsidiarity and the City', pp. 68-72 (n 13). 46 https:// www. un. org/ en/ desa/ around-25-billi on-more-people-will-be-living-cities-2050-proje cts-newun-report (last accessed on 12 March 2023). 47 See, e.g., Bogdanor (2019), p. 171. it, like the Irish nation did in 1921. 48 As a result, the nation of England is one of the most centralized countries on the European continent. 49 Among the ten largest urban areas in the UK, nine are located in England. 50 Yet only Greater London has a distinct constitutional status. In itself this is not surprising. London is the most densely populated city, the only city in the both countries studied in this paper that comes close to a megacity and the UK's capital. In 2000 the Blair government established the Greater London Authority (GLA) as a result of a 1998 local referendum resulting in the Greater London Authority Act. 51 On that moment, a period of fourteen years during which the city had no elected metropolitan government at all came to an end.
The GLA differs in various ways from other local government institutions in the UK. 52 It is a jurisdiction of its own governing over 9 million people. It comprises a mayor, an assembly and a small administration, as well as 33 borough councils that are responsible for the various districts in the city. The major and the assembly are elected by the Londoners through an innovative and proportional voting system, making them the only elections in the UK that utilize this type of system. The powers of the GLA are diverse. 53 It directly handles the London fire commissioner and strategic housing funding. In this sense, the GLA occupies a distinct position in the UK's local government landscape. Yet the GLA remains significantly dependent on the decisions of the Westminster Parliament. Although the GLA has full executive power over particular London services such as Transport for London (TfL), the largest part of its funding comes from the central government. 54 The influence of the GLA on policy areas that are typically important to the city context, such as health, education and social care is also limited for the same reason. 55 The GLA could therefore be constitutionally positioned between the entrenched, devolved systems that are in place in Scotland, Wales, and Northern Ireland on the one hand, and the traditional system of local government in England on the other hand.
The amount or the significance of the powers that the central government attributes to the city level does not necessarily paint a complete picture of what autonomy could entail for urban areas. Autonomy is an inherently unstable social construct, being very much dependent on the specific circumstances as to how it is pursued. 56 In this perspective, cities can play their part in shaping their own autonomy 48 Bogdanor,Beyond Brexit,p. 199 (n 25). 49 Kenny et al. (2018) 54 Greater London Authority Act 1999, section 31(1)(a). 55 Greater London Authority Act 1999, section 31(3). 56 Bulkeley et al. (2018) at 705-707. by developing their own policy goals in collaboration with the private sector and NGOs. The GLA has, for instance, developed its own sustainability agenda with specific policy goals in areas such as transport, energy and climate change. 57 By seeking collaboration with both public organizations, such as London First and the City of London Corporation, and local private communities, such as the Brixton Transition Town movement (BTT), the GLA has managed to adopt a 'low carbon zone' in the city. 58 This development illustrates that local institutions on the local level can generate their own autonomous position in tackling particular issues, even if they are not the main responsible actor from a constitutional perspective.
The approach to autonomy as being circumstantial is especially appealing to the context of cities of which the exact boundaries are equally unstable and shifting. If a city is able to tackle particular issues without the need for intervention by the central government, it generates autonomy which could justify the attribution of even more powers to the city level. Initiatives such as the establishment of the sustainability agenda in London illustrate that the attribution of even a relatively small number of powers may form a sufficiently stable basis for the development of a distinctive unit based on social solidarity. The conclusion seems to be warranted that the Greater London area is able to handle additional devolved powers. 59 Then, the question should also be asked whether Greater London is also distinctive enough from the rest of the country to reach this conclusion. This is a matter of speculation. It could safely be said that London is an outlier, being a 'global city' 60 that has for centuries been much more diverse than any other region in England or the United Kingdom. 61 The 2016 'Brexit' referendum has stressed the divide between 'metropolitan values' in the larger urban areas 62 on the one hand and those of almost anywhere else in England on the other hand. 63 After the referendum, approximately 175.000 Londoners signed a petition inciting mayor Sadiq Khan to secede London from the UK to keep the city in the EU. Khan did not directly endorse this idea, but in 2016 he called for a 'new constitutional deal' involving the attribution of new powers to London. 64 Significant changes to London's constitutional position are unlikely to happen at short notice. In the long run, however, this might change. 57 Bulkeley et al. (2018) at 705-707.
The position of other English urban areas is rather different than that of Greater London. Although their boundaries are equally shifting and circumstantial, they have had more difficulty in attaining an autonomous position. The enactment of the Local Democracy, Economic Development and Construction Act 2009 has introduced the so-called 'combined authority', allowing a group of local authorities to receive some delegated powers from the central government. This model, firstly adopted in Manchester, the UK's second-largest city, involved the introduction of a modest form of devolution. It holds that an association of local authorities take over responsibilities for specific issues, such as urban regeneration, economic development and urban growth, which none of these authorities could handle on their own. The newly established organization in this framework is the so-called combined authority. The foundation of the Greater Manchester Combined Authority in 2011 was followed by nine others. 65 The introduction of this modest model of devolution has not proved to be very effective in any of the urban contexts in which it was adopted. Not only does the operation of these combined authorities appear to be very dependent on the central government, it also has the inherent weakness that it leads to competition among metropolitan and urban areas, especially those that are relatively close to each other. As Parr notes, 'it is possible that combined authorities in metropolitan areas, motivated by local pride, may embark on grandiose infrastructure projects such as a sport arenas, concert halls exhibition centers, etc. These would, no doubt, be seen by local leaders as "worthy of the area" and also likely to be accompanied by additional streams of economic benefits'. 66 This leads to the tentative conclusion that, for the time being, the English urban areas apart from Greater London, do not as such appear to be sufficiently distinctive in order to be granted distinctive autonomous powers.

The Potential for the Recognition of Cities in the Dutch Constitution
In the constitutional structure of Dutch local government, cities do not really have a distinct position. The Dutch Constitution only recognizes municipalities as such without distinguishing between large cities including Amsterdam (903,399 inhabitants) and small rural areas such as the island of Schiermonnikoog (944 inhabitants). 67 That does not come as a surprise, given the high degree of urbanization in the country. Yet, Amsterdam, the largest city of the country (903.399 inhabitants), 68 does not feature on the list of the 200 largest cities in the world. 69 The Dutch sociologist Abram de Swaan argued already in 1991 that the Netherlands could, given its high degree of urbanization, be regarded as a very large conurbation rather than as a country. For De Swaan, the largest part of the country consists of the 'Rim City' (Randstad), consisting of the four largest cities of the country: Amsterdam, Rotterdam, The Hague and Utrecht. It starts with a pole in Eindhoven ranging via Breda and Dordrecht (in the south of the country) to Rotterdam, Delft and The Hague (in the southwest). The strip of urban areas continues to the northwestern part with cities as Leiden and Haarlem leading to Zaanstad, Amsterdam, Het Gooi and Utrecht. The total length between the two poles is approximately 175 km and breadthways 15 km. 70 The Rim City is the fourth-largest metropolitan area in Europa after London, Paris and the Rhine-Ruhr area in Germany. Yet it cannot be properly regarded as a megacity for the purposes of this paper as it does not, in contrast to London or Paris, generally operate as a unity. In this region, the only players that are being constitutionally recognized in local government are the respective municipalities. If this traditional view on local government were discarded, it would however not be an exaggeration to describe the Rim City as a megacity. If the Rim City's surface area were to be compared with that of Greater London, it would even seem reasonable to include somewhat more remote urban areas, such as Zwolle, Apeldoorn, Arnhem and Nijmegen (in the east), within its boundaries, constituting a population of more than 8 million inhabitants. 71 Only a sparse amount of mostly rural areas, covering the provinces of Limburg, Friesland, Drenthe and Groningen and the region of Twente, would then fall outside the borders of this metropolitan area. Even these areas include important cities of more than 100.000 inhabitants such as the city of Groningen, Enschede and Maastricht. From this perspective, there does not seem to be a need for intra-national devolution settlements such as those in the UK. It is not very likely that this will change in the near future, as the rigid structure of the Dutch Constitution makes it very hard to substantially revise it. This most certainly also applies to the structure of local government, as regards to which many unsuccessful constitutional reforms have been proposed. 72 Similarly to cities in the UK, the boundaries of Dutch urban areas seem to shift. Two interesting developments should be noted in this regard. First, the municipality has already for many decades been the most important tier of government at the expense of the role of the provinces, the latter having become mainly responsible for administrative supervision on behalf of the central government and fulfilling mainly some autonomous tasks in the spatial domain. The importance of the role of municipalities has been reinforced by the recent decentralization programmes in the fields of social care and spatial planning that have been in effect since 2015. As Groenleer and Hendriks point out, these programmes have resulted in an increased collaboration between municipalities on both a mandatory and a voluntary basis. 73 This may eventually result in the formation of new urbanized regions which are centered around one or more cities, such as Enschede in the region of Twente or Arnhem and Nijmegen in the city region of the same name, both in the eastern part of the Netherlands. 74 Given the relatively large surface area of these regions, this could further strengthen the ties between relatively small rural municipalities and larger urban municipalities thereby increasing their importance in relation to the central government.
The second development is the emergence of the so-called 'emergency response council' (Veiligheidsberaad). When the Covid-19 pandemic hit the Netherlands early 2020, there was no proper law in effect on civil contingencies. Instead, as a temporary solution, the central government granted far-reaching powers to the so-called 'emergency response council' consisting of the 25 presidents of the 'emergency response regions' (veiligheidsregio's). Each emergency response region is a group of municipalities aimed to co-operatively manage acute emergencies, of which the formation was required by law. 75 The president of the emergency response region is, by definition, the mayor of one of the most important municipalities within that region, appointed by the government. 76 Before Covid-19 entered the Netherlands, the emergency response council was rather dormant, but this changed soon after. During the first months of the Covid-19 pandemic (March-December 2020), this council has taken far-reaching measures in order to combat the pandemic, partly on the advice of the Dutch central government ministers of Health and Justice. After 2020, the emergency response council has remained in charge, albeit more on the back seat. It is however still too early to tell if the institution of the emergency response council and the respective emergency response regions will remain as important as it now seems. Similar to the increased collaboration between larger and smaller municipalities, the collaboration between municipalities in emergency response regions may even further blur the division between the rural and urban areas throughout the country.

The Multifacetedness of Identity: Spatial Identities as a Model
The constitutions of the UK and the Netherlands, constituting unitary states at least in a formal sense, leave relatively little room for the attribution of specific authorities to urban regions. Moreover, it is relatively hard to demarcate the boundaries of urban areas that are so distinctive that they deserve constitutional recognition. 73 Groenleer and Hendriks (2020) at 202-203. 74  Therefore, the structure of local government in both constitutions should be developed in an informal way. In this section, I argue that the absence of clear boundaries of urban areas does not necessarily have to pose a problem. First, I explore the distinction between thick (traditional, deeply rooted) and thin (transitory, mainly economical) identities (Sect. 4.1), which help to understand how regional identities may shift. Secondly, I aim to show that these different forms of spatial identity could be the basis of a new constitutional narrative, which is tailored to the informal way in which the structure of local government develops in both the Netherlands and the UK (Sect. 4.2).

Thick and Thin Identities
Traditionally, identities are linked to the context of the nation state. Just as the development of a constitution, it takes a long time for a nation to develop its own distinct identity. In a similar way, other spatial identities, for instance linked to regions, may develop.
The Finnish geographer Anssi Paasi has distinguished four stages in which a region may come to rise, which do not necessarily need to follow each other and which may even happen simultaneously. 77 First, there is the development of the territorial shape of a region, which can be the result of either a long historical process or, coincidentally, a decision by the responsible authority. This territorial dimension helps to distinguish the respective regions within a state, which may be relevant in order to determine what region may receive financial support from the central (or EU) government. Second, the emergence of the symbolic dimension refers to the extent in which specific symbols, such as a flag or a (regional) language, come to rise which underline the existence of a specific region. The third institutional stage is related to the territorial and the symbolic dimension of a region. The emergence of both the territory and one or more symbols in a region may trigger the establishment of regional institutions that can be seen as an expression of the distinctness of a specific region which gives room for 'regional ways of doing things'. This institutional dimension does not only refer to the mere existence of institutions for the region, such as local authorities and companies, but also to their interaction with others inside and outside the region. Fourth, the functional dimension refers to the extent in which a region is externally recognized as such in larger systems, for instance by neighboring regions, the central government, or even supranational institutions.
All these dimensions of the development of regions are interrelated and may both internally and externally transform in the course of time. Thus, Paassi refers to regions as 'a perpetual and dynamic process of scaling the practices and discourses through which the previous shapes are produced and reproduced'. 78 In these processes, a distinction can be made between regions with a thick (historically and institutionally embedded) and thin (transitory and future oriented) identity. Generally, regions with a long history and with a great deal of shared, stable, institutions have a 'thick' identity, while the identity of regions without these characteristics can be described as 'thin'.
Interestingly, the political geographer Kees Terlouw has observed that globalization processes have highly influenced the emergence of regional identities in various ways. 79 First, these processes have generally undermined the position of nationstates, as regions are better equipped to serve the needs of international companies in a global context. This has strengthened the position of the most competitive regions, while that of many peripheral regions weakens. Second, the regional identity of regions with distinct language or cultural rights, such as Scotland, has always been thick on the basis of disagreement with the policies of the national government. Third, the functional shape of regions has become more susceptible to change, as regions increasingly need to adapt to the changing needs of a globalized world. This has forced the establishment of new regions by changing the administrative borders, which generally undermine existing (thick) identities. Moreover, the susceptibility of newly established regions to change makes it harder for them to fully develop a new thick identity. 80 Of course, it is not possible to make a sharp distinction between 'thick' and 'thin' identities, as in many regions these two types of identities come together. 81 For instance, relatively modern regions such as the Rim City may try to make their 'thin' identity more thick by referring to their glorious past which took place before the cities in this region started to co-operate. Conversely, traditionally established regions, such as Scotland, try to establish a thin, future oriented, identity by referring to their original, thick, identity which stems from their shared history.

Identities and the Urban-Rural Divide
The concept of spatial identities helps to better understand the way in which the urban-rural divide could be bridged in the UK and the Netherlands from a constitutional law perspective. Constitutions have traditionally concentrated on institutionalizing collective autonomy in representative bodies on both the national and local level (parliaments and local councils) which are based on social solidarity. Given the potential diverging interests between urban and rural areas, it is not always easy to discern what this notion of social solidarity precisely entails. The concept of spatial identities could help to better understand how the public interest should be defined and, more importantly, how this may render the members of society willing to accept decisions that may be beneficial to the common good despite being harmful to their own self-interest. Three aspects could be distinguished here.
First, it is important to reiterate that spatial identities are inherently unsettled, thereby tailored at continuously shifting borders within a country. This is especially beneficial to the structure of local government in the constitutions of the 79 Terlouw (2009)  UK and the Netherlands, since they always have been subject to organic evolution. If the concept of spatial identities is applied to the contexts of both countries, there is no direct need to entrench the rights of specific spatial areas within the country as their borders may change relatively easily. Thus, from a constitutional design perspective, it may be even beneficial that the respective constitutions of the Netherlands and the UK (including its respective constituent parts) have never had a strong tradition of granting much autonomy to urban areas. Otherwise, if new (city) regions with a 'thin' identity emerge, constitutional law makers would have to make the complicated decision under what conditions they can be seen as an urban region and when new powers should be granted to them. Hence, the concept of spatial identities helps to appreciate the bottom-up character of the constitutions in both countries.
Second, the concept of spatial identities also helps to see the interrelatedness between urban and rural areas. This applies particularly to the emergence of clusters of collaborating municipalities and the so-called emergency response region in the Netherlands, and the so-called combined authorities in the UK. In some cases, as in the Arnhem-Nijmegen region in the Netherlands, new urban regions could emerge in places that were originally considered to be rural. Although the regional identities of these clusters of local authorities will initially, at best, be of a thin, future-proof, character, some of them may attain a 'thicker' character if the citizens involved appear to benefit from it. As a result, in unitary states such as the Netherlands and the UK, there is no need to make a sharp distinction between urban and rural interests. This would in general be a reason to not constitutionally entrench the position of cities in both countries.
Third, the distinction between thick and thin identities may also help to appreciate why in some exceptional cases there may be a reason to grant a greater level of self-government to the economically most important urban areas in a country, such as Greater London. The economic success of the UK's capital benefits the whole country as it attracts tourists, investors, and generally talented people. Moreover, London itself already has a 'thick' identity, which may help to prevent the complications that may arise when having to demarcate the borders of this urban area. Admittedly, there is a relatively great divide between London's identity and the perceptions about being British or English in the rest of the country. However, by stressing London's importance for the rest of the country, a 'thin' identity could be created which could justify the decision to further strengthen the constitutional position of this urban area.

Concluding Remarks
The urban age is undoubtedly challenging for constitutional scholars. Cities appear to have an enormous potential for tackling global challenges. It is however not easy to convert this new reality to specific constitutional contexts. This appears to be especially true for countries in western Europe, such as the Netherlands and the UK. Both countries are largely urbanized and established by constitutions which place high hurdles for formally entrenching the constitutional position of cities. Therefore, if the position of the city is to be constitutionally recognized, it should be done so in an informal way for instance by using devolution arrangements.
Even if this is accepted, this is easier said than done. The requirements of the democracy requirement inherent to the rule of law are not easily met. Adherents of city autonomy usually justify the constitutional recognition of cities in terms of subsidiarity, but this argument places a high burden of proof on adherents of city autonomy. First, they would need to prove that cities are better capable than other levels of government to tackle particular policy issues. Second, they should show that a greater autonomy of cities does not distort the balance of interests between urban and rural interests.
The latter requirement appears to be especially problematic in the Netherlands and the UK. The blurry division between cities and non-cities in a largely urbanized context makes it hard to attribute powers to cities in a way which could make sense to citizens in both these cities and non-cities. Only London, being a global city that is more diverse than the rest of the UK, is unique in this way. The case of London shows that powers do not necessarily have to be directly attributed from the central government. Rather, cities may be able to generate an autonomous position by developing their own policy goals, for instance in collaboration with the private sector and NGOs. The development of particular initiatives, such as the GLA's sustainability agenda could also help to make the case for devolving more powers to the city level from the central government. This shows that being a distinctive urban unit could help to make the case for more autonomy. Autonomy is thus not granted, but could also evolve.
The context of other city regions in the Netherlands and the UK seems to be more indeterminate. In these highly urbanized countries, it is hard to discern parts of the countries that are truly rural in the sense that they are not impacted by urbanized demands. The emergence of the emergency response regions in the Netherlands and, to some extent, the creation of combined authorities in England, show how rural areas are increasingly falling within the reach of urban areas. The idea of differentiating between urban and rural areas therefore seems to be problematic. The model of spatial identities, which consists of thick (traditionally, historically, and institutionally embedded) and thin (transitory, future oriented) identities helps to explain this. This model not only reveals that the exact borders of urban areas are continuously susceptible to change, but also that the co-operation between urban and rural local authorities could lead to the emergence of new urban regions. Hence, there is no general urgency to constitutionally entrench the position of cities in these constitutional systems, with Greater London as the only possible exception.
The politically enforced constitutions of the UK and the Netherlands are usually dependent on organic developments rather than of deliberate constitutional design. The most important criterion for constitutional change has been delivered by J.A.G. Griffith. He stated: 'If it works, it is constitutional'. 82 Both constitutions have, to date, appeared to work, in the sense that there has never been substantial protest against the way in which they are structured. The constitutional position of cities therefore should be mainly left to organic developments rather than of deliberate constitutional design.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.