The Views of Neighbourhood Associations on Collaborative Urban Governance in Tallinn , Estonia

Throughout the 21st century a clear shift from hierarchical government towards network-like governance is evident in Europe and countries of the former Soviet Union. A door has been opened for citizens to assert their social and political citizenship and for the public sector to systematically search for novel ways to improve and expand representative practices. Estonia may be considered a post-socialist ‘market-experiment’, as almost no urban land belongs to the public sector. Thus, all decisions related to urban spatial changes are strongly influenced by private interests. The situation of ad-hoc planning, which intensified during the real-estate boom of the mid-2000s, evoked the mushrooming of civil activism, mostly in the form of neighbourhood associations. Current research investigates the roles and efficiency of neighbourhood associations in the collaborative urban governance of Tallinn. The study shows that neighbourhood associations help to reinforce people’s social and political citizenship; however, there are still insufficiencies regarding their ability to represent different interests in Tallinn and in their own neighbourhood. We also discovered that Tallinn, as a post-socialist city, is currently in an experimental phase of learning how to efficiently gain from collaborative networks in the urban governance processes.


Introduction
Recent years have shown a growing interest in the intersection of social justice and urban diversity and developing innovative approaches to urban governance and planning (Taşan-Kok et al, 2013).Boundaries between public policies, market actors' activities, and the activism of civil society organisations are increasingly becoming restructured towards close cooperation between entities.This is consistent with the wider shift in all governmental levels from hierarchical government models towards horizontal network-like governance practices (Gaventa, 2004;García, 2006;Häikiö, 2007;Martínez, 2011).
In this article we focus on one particular type of entity in urban governance-the neighbourhood association (henceforward NA)-by showing what is and what could potentially be its role in urban governance.NAs as mediators for citizens' voices (Weare et al, 2009;Jun and Musso, 2013) aim to reveal public interests and communicate them efficiently to an administrative authority (Ansell and Gash, 2008).Such grassroots initiatives, valuable partners for public bodies, have sometimes been nurtured by specific support mechanisms (Jun and Musso, 2013).As urban spatial changes shape immediate urban environments that citizens experience, it seems that cooperation during urban planning processes is especially important.Therefore, NAs often focus their activism towards sociospatial development related issues like land development and public order in the neighbourhood.It is assumed that a collaborative atmosphere, including a sound cooperation between public bodies, market actors, and citizens' organisations, helps to make sense together through the power of better argument (Healey, 1997).A collaborative atmosphere also increases the democratic legitimacy of sometimes remote and non-transparent decisionmaking processes related to changes in urban environments (Connelly, 2011).
According to Raagmaa and Stead (2014), post-socialist planning systems hitherto contain some relics of hierarchical (top-down) Soviet planning but despite this, all cities in countries formerly under central planning and communist regimes have gone through an enormous change in their planning and urban governance traditions.We study the role of NAs in urban governance in a post-socialist context in Tallinn, the capital city of Estonia.Collaborative governance and urban planning principles, taken as exemplars from the legislation of Western and Northern European countries, were officially incorporated into Estonia's legislation almost two decades ago.Yet, the institutionalisation of these principles-making collaborative governance and urban planning principles take root-has taken more time.
In our article we are focused on two issues.First, we are interested in what can be learned from the experience of Tallinn regarding the role of NAs in urban governance in general.Initial observations of our study showed that the rise of new NAs in Tallinn coincided with the economic boom years of the mid-2000s (Tammaru et al, 2009;Kährik et al, 2012).It was quite evident that the increasing citizen activism served as a response to the period when larger spatial restructuring in urban areas was expected.We explore, through empirical analysis, the roles of NAs in such situations.The results contribute to theoretical debates of urban governance and planning that focus on the roles of various urban actors in situations of remarkable urban spatial transformations.
Second, neighbourhood activism in urban issues is still a relatively new phenomenon in Estonia as well as in other post-socialist countries (compared to older democracies in Europe).We wish to understand how rather normative principles of collaborative urban governance and planning are translated to a post-socialist context.Although the planning legislation and official legal principles for citizens' rights are quite similar in traditional democratic societies and in post-socialist democracies (the legislation is sometimes even adopted verbatim), the wider context for political culture is different.For example, with the ownership reforms of the 1990s, the former full state ownership of urban land and properties was transformed to a full-market situation (Lux et al, 2012).Contemporary Estonia is sometimes considered to be a true 'market-experiment' (Tammaru et al, 2015) in which virtually no urban land is publicly owned.This is not the case in most of the traditional market economy-oriented countries of Europe.In Estonian urban planning, the concurrent phenomenon tends to be ad hoc planning culture dominated by splintered private sector interests in the context of an ultra-liberal economy (Ruoppila, 2007;Leetmaa et al, 2009;Raagmaa and Stead, 2014).In this situation the grassroots reactions seem to be a counterbalance for the weak capacity of the public sector to coordinate the accompanying remarkable spatial effects.

Neighbourhoods and neighbourhoods associations in urban governance
The discussions on collaborative urban governance and planning have lasted for a long time in literature and include many sub-topics (Healey, 1997;Fainstein, 2000;Innes and Booher, 2004;Ansell and Gash, 2008;Healey, 2015).To understand the role of neighbourhood associations in collaborative urban governance, the 'just city' discussions are relevant and informative.For Fainstein (2000), inclusive, empowered and just city means a situation where the inhabitants not only have a formal say in decision-making processes, for example in planning issues, but they indeed have the opportunity to contribute to the real outcome of their immediate living environment.
The obstacles for an inclusive decision-making process mostly originate from two strands.First, the spatial planning process is frequently characterised by the need to have (technical) expert-level knowledge and skills to be able to effectively influence the physical interventions made in the city or also in one's own neighbourhood.Consequently, the decisions related to urban changes are often distant from the residents and other actors who actually live and work in this environment later on.Individual voices working in isolation are usually unable to survive in this complex environment, but citizens gathering in NAs operating as part of a collaborative team could potentially benefit from each other's professionalism and become accountable partners for professional planners and public bodies as well.
Second, a representative democracy in an urban decision-making process is often not combined sufficiently with other supporting participative democratic mechanisms.Although planning procedures initiated and carried out by government bodies might seem distant and too technical, they possess legitimacy and accountability obtained through constitutional elections (Plotke, 1997;Häikiö, 2007;Davoudi and Cowie, 2013).The issue is how representative practices can be improved and expanded.The challenge for collaborative urban governance is to include more interests and voices without losing democracy (Connelly, 2011).Various forms of citizen input to planning discussions regarding their living environment make the discussions often more complex and less direct (Plotke, 1997), but at the same time these discussions inform public decision-makers about the diverse preferences in a complex urban society.The attention or curiosity the civil organisations take notice of in their immediate surroundings potentially brings about improvement in otherwise uniform public services.In some cases, various citizens groups help to innovate in public service delivery by pointing to the needs of specific population groups or to expectations about how something could be better organised for neighbourhoods (García, 2006;Martínez, 2011).However, neighbourhood representatives often have to very thoroughly defend their inherent democracy and representativeness.For example, they need to demonstrate transparently who and what interests they actually represent, because in some cases the NAs only extend representational democracy, excluding vulnerable groups who do not gather in organisations so easily or whose voices are weaker.Furthermore, based on Purdue (2001), community activism tends to rely 'on a network structure and often lacks the clearly defined institutions of power and legitimacy of a political party ' (p. 2214).Yet, according to Plotke (1997, p. 24), 'rather than opposing participation to representation, we should try to improve and expand representative practices' for recognising innovative solutions to urban issues and for implementing the 'social and political citizenship' (García, 2006, pp. 748, 750) of urban residents.Fincher and Iveson (2008) conceptualise social citizenship by tying the goals of urban governance and planning to social aspects of redistribution, recognition, and encounter.They argue that extreme poverty and deprivation also produce extreme exclusion of some population groups from societal life.Redistribution (ibid., is always tied to spatial decisions-some planning outcomes reduce disadvantages and inequality and others may reinforce it.A diverse set of ambitions, preferences, and endeavours that meet in the contemporary city require actors to increasingly listen to others while preserving a certain curiosity about various voices.Therefore, Fincher and Iveson (2008) differentiate two approaches for recognition-pragmatic and contextual (p.104).With the pragmatic approach, an affirmative checklist method is used when recognising certain groups, often disregarding embedded differences.The contextual approach, on the other hand, is more of a relational method, where recognition is strongly entwined with the concept of intersectionality (Valentine, 2007).This more sensitive way of recognising different groups takes into account all social roles people may carry simultaneously and acknowledges that the set of voices in a city is dynamic (since new groups may always emerge).Thirdly, Fincher and Iveson (2008) argue that spatial decisions should provide places of encounter (p.151-159) for various recognised groups so that people can meet others alike or make contact with the new and different, helping to overcome social distances that threaten contemporary urban life.
To implement political citizenship, an institutional context is needed that acknowledges participative mechanisms side-by-side with representational ones, and also provides respective mechanisms for citizen participation (García, 2006).Furthermore, Gaventa (2004) argues that more empowered forms of participation in local governance can lead to democracy-building and even pro-poor development outcomes.For this reason, it is important to carefully observe 'how the spaces for participatory governance work, for whom, and with what social justice outcomes' (ibid., p. 31), meaning whose voices are really heard and whose are excluded (ibid., pp. 27, 38).
These discussions are worth reconsidering when making arguments about the role of neighbourhood activism.Regarding the issues of participatory democracy in urban governance, neighbourhoods and their representative bodies are in the centre of discussion in a range of research (Purdue, 2001;Häikiö, 2007;Lowndes and Sullivan, 2008;Connelly, 2011;Häikiö, 2012;Pill and Bailey, 2012).According to Lowndes and Sullivan (2008), a neighbourhood is an appropriate unit for urban governance for many reasons.For example, a neighbourhood contains fewer citizens, thus making direct participation feasible; also, neighbourhoods are expected 'to encapsulate homogenous communities with shared values, beliefs and goals ' (ibid., p. 57).As such, due to the force of homophily (McPherson et al, 2001;Weare et al, 2009), neighbourhoods bond together certain groups of people based on shared ethnicity, lifestyle, or preferences.
As a counter-argument, there is rich evidence that more and more people in contemporary cities live in diverse rather than in homogeneous neighbourhoods (Holloway et al, 2012).This again raises the question of appropriate representation.When people with very different socio-economic and cultural backgrounds share the living environment, the visions of these groups concerning the neighbourhood's future are not necessarily similar.Furthermore, neighbourhoods are often in transition rather than socially stable in terms of residential mobility.For example, inner-city neighbourhoods undergoing gentrification attract younger, more active, and often economically better established social groups like young entrepreneurs.These citizens have obtained certain socio-economic status and may be more willing to invest their time in neighbourhood development as well (Jun and Musso, 2013).However, if the neighbourhoods are socially diverse, because the 'old' not-so-active residents still live here as well, but the NAs rather deal with topics important for newcomers, the question whether these NAs are able to represent the whole neighbourhood should be addressed.
Gaining legitimacy is a discursive process dependent on context, shared beliefs, and cultural values, which reflect different positions of stakeholders in governance networks (Häikiö, 2007;Häikiö, 2012).We know that the leaders of NAs often aim to represent the collective interests of a neighbourhood (Purdue, 2001).However, the legitimacy of their positions is related to their track record of delivered outcomes or competence in a given issue, and personal charisma and trustworthiness (Connelly, 2011).Therefore, legitimacy is constructed both through formal and informal processes (ibid.;Davoudi and Cowie, 2013).Accordingly, all the abovementioned questions about how NAs are able to reinforce local democracy and support social and political citizenship should be posed if the aim is to understand the potential role of neighbourhood associations in urban governance.

Socio-spatial changes and urban governance practices in post-socialist Tallinn
We study the role of NAs-non-profit organisations driven by citizen initiative and mainly committed to preserving or improving neighbourhood qualities-in a post-socialist European city, Tallinn (approx.400 000 inhabitants), the capital of Estonia.Much like in other Central and Eastern European countries, the urban planning institution in Estonia has undergone a remarkable transformation in the context of vast socio-spatial changes (Liepa-Zemeša and Hess, 2016).Among other post-socialist European capitals, Tallinn is characterised as a city where social inequalities and the resulting socio-economic segregation have rapidly increased in post-socialist decades (Tammaru et al, 2015).
Tallinn is composed of a mixture of diverse urban districts originating from various historical periods.Half of the city's buildings were demolished in WWII and the remaining historical housing (mostly 1 to 3-storey wooden houses) survived relatively unscathed (Ruoppila, 2007).This is in line with the trajectories of inner cities elsewhere in Central and Eastern Europe (Kovacs et al, 2013), where during the socialist years the pre-War residential quarters were left to decay (Hess and Hiob, 2014).The main efforts and investments of housing policies were directed towards the high-rise panel-housing estates (Kährik and Tammaru, 2010).In the context of severe housing shortages in growing socialist industrial cities, an apartment in a panel-housing district was a sought-after residential solution.Today, approximately 2/3 of the population in Tallinn lives in these housing estates.
Over time, a shift in residential preferences has occurred.The apartments in housing estates gradually have lost their prestige and older inner-city districts have become attractive gentrified places (Kährik et al, 2016).Although the panel-housing areas have maintained a certain mixed socio-economic status (Kährik and Tammaru, 2010;Temelová et al, 2011;Marcińczak et al, 2015), gradual sorting of high social status people from the socialist modernist housing and into suburbs or gentrifying districts is an ongoing reality.The former inner-city districts have transformed into diverse social environments, where both former residents (often less affluent, older, and with industrial worker background) live alongside newcomers (young households, students or specific lifestyle groups who value culturally attractive historical neighbourhoods near the amenities of the city centre).Interestingly, these hot spots of gentrification also offer fertile ground for local activism (Leetmaa et al, 2015b).
In addition, the socio-spatial transformations in Tallinn contain an ethnic component.Slightly less than half of the population of Tallinn is Russian-speakers, or their descendants, who arrived in the country during the Soviet decades.As a result, Tallinn has a dual ethnic character-ethnic Estonians and Russian-speakers live somewhat parallel lives with only limited intergroup communication (Kamenik et al, 2015;Leetmaa et al, 2015a).Russianspeakers, usually skilled-workers with an industrial background, who do not speak Estonian well, have found themselves in a more disadvantaged position in the growing serviceoriented labour market (Leping and Toomet, 2008).They change their place of residence less often and if they do, they have a tendency to select destinations where familiar culture networks exist (Mägi et al, 2016).Estonians, on the other hand, are leaving the panelhousing districts for inner-city neighbourhoods more frequently.Therefore, in panel-housing districts, the proportion of Russian-speakers, which in general is 'less connected' with the labour market and with societal life in general, is relatively higher; similarly, in gentrified districts, the long-term population with a relatively lower social status tends to be the Russian-speaking sub-population.Consequently, the ability to perform one's social and political citizenship is higher among ethnic Estonians in Tallinn as they form the majority of inner-city residents where neighbourhood activism is more visible.
Traditionally, in a European city, significant public resources are invested to prevent the market from dictating spatial and social development (Ruoppila, 2007).There are also significant differences in the role of the welfare state among European countries (Kazepov, 2005), which also applies to planning systems: the welfare-systems of Central and Eastern Europe are typically weak.Raagmaa and Stead (2014) argue that 'spatial planning in Central and Eastern Europe differs from Western and Northern Europe in terms of rapidly changing economic, organisational and political landscapes, lower levels of trust in the role of government, and the position of planning in society' (p.672).Europeanisation-attempts to apply certain elements of western models of planning-has been recognisable but in comparison to the rapid change of other institutions, the replacement of old planning traditions with new practices has been a somewhat slower process (Raagmaa and Kroon, 2005;Raagmaa and Stead, 2014).An explanation for this delay lies in the fact that urban housing stock and land became almost entirely privatised in the 1990s.Consequently, any (urban) planning action today is greatly influenced by the interests of private owners.Ruoppila (2007) admits that a generally liberal approach towards urban development prevailed in the 1990s (see also Leetmaa et al, 2009).He argues, however, that from the late 1990s onward, the role of planning has gradually strengthened: purposeful steps were taken to integrate physical planning and real-estate regulations, to increase transparency in planning and city management, to achieve greater involvement of the general public in planning processes, and to promote sustainability.Thus, collaborative practices in urban governance also began to become more common only in the 2000s, suggesting that new urban governance practices require time to get rooted.
An incentive for change in governance practices was also the nascent recovery of the nongovernmental sector.According to Ruutsoo (2002), the 1990s did not contribute much to the emergence of Estonian civil society as the state administration had not yet considered civil society as a resource.This very poor understanding of civil society as a potential state partner resulted in an elitist governance orientation (ibid.).Based on Rikmann et al. (2010), the state consciously started to shape the operating environment of the NGO sector in the 2000s.In line with the Estonian Ministry of the Interior1 , the recognition that citizens' associations need consistent support was achieved, and many state-budget-funded services began to be provided for NGOs, for example counselling, various mentoring and development programmes, project-based-funding, access to research and analyses, etc.This has resulted in the professionalisation of some of the organisations via recognition and stable funding.Yet, at the same time, this professionalisation has increased the threat that pre-existing collaborations exclude those who do not have personal contacts with decisionmakers (ibid.), or who just might 'not have the time, the capacity, the know-how, or the political resources to participate' (Davoudi and Cowie, 2013: p. 564; see also Rikmann et al, 2010).

Description of the fieldwork
There were twenty two NAs in Tallinn as of May 2014.We observed that NAs mostly existed in four city districts 2 -Põhja-Tallinn (Northern Tallinn), Nõmme, Kesklinn (City Centre) and Pirita (see Figure 1)-with the exception of one in Haabersti.Most importantly, we observed that NAs are intrinsic to historical inner-city areas (City Centre and Northern Tallinn) and lowrise pre-WWII districts with single-family houses (Pirita and Nõmme).There were, however, no NAs in city districts with socialist-era panel-housing estates.
Due to the relatively low number of NAs we attempted to interview at least one representative from each.We were able to make contact with fourteen NAs out of a total of twenty-two.We asked for a representative of an NA who would be able to discuss the following issues: how and why the association was founded; what the main activities are; who they represent; and how they perceive the collaborative practices in Tallinn.Based on some background information on the recent activities of the NAs, we are able to say that the remaining eight NAs, which did not respond to our endeavours to meet, were also less active or not yet fully established.Therefore, fourteen qualitative expert interviews with founding members and/or current members of the board of directors of NAs were conducted.We must stress that certain views presented in our results may reflect the views of the representatives of the NAs and not the whole association, not to mention the whole neighbourhood.Therefore, we aimed at being as aware and critical as possible: we fully acknowledged that some views or interests of other active NA members might be left out.We also understand that some presented views might be slightly overcritical, as we have not made a specific survey among other urban actors (developers, public representatives, etc.) to reveal how the latter might perceive the activities and the legitimacy of NAs.However, the leaders of NAs are usually in direct contact with the local government and other stakeholders, thus having enough experience and professional expertise to explain from their point of view the power of neighbourhood activism in the city.Table 1 gives an overview of the NAs, their goals and the métier of the interviewees.The interviews were conducted between February and May 2014.The lengths of the interviews varied from forty-one minutes to one hour and twenty-two minutes.After the first round of analysis we lacked sufficient understanding of why neighbourhood activism is less frequent in panel-housing districts and among Russian-speakers.We therefore conducted an additional interview in June 2015 with the representative of the Lasnaidea initiative (thus, in total we carried out fifteen interviews, see Table 1).
The main interview questions addressed issues of representation, the interest groups in the neighbourhood, the level of competence of various stakeholders, and the overall situation of collaboration in urban governance and planning.Although we followed a semi-structured interview guide, we included additional issues that arose spontaneously during the interview.The empirical analysis followed the principles of directed content analysis (Hsieh and Shannon, 2005), meaning that the main codes for analysis were drawn from theory.Some coding examples include representation; interest groups in planning and urban governance; and incentive for collaboration.This was supplemented by open coding, meaning codes were also derived from data and the researcher's knowledge of context: the leaders of NA; neighbourhood activities; and concerns and evaluations regarding collaborative practices.The main results-the emergent pattern of NAs, who they represent, and what are their roles in Tallinn's urban governance-are presented in the following section.
Open Access Journal 57

The Mosaic of NAs
Altogether an interesting set of NAs was obtained, reflecting the four mentioned districts of Tallinn quite well.For example, there are five rather different NAs within Northern Tallinn, where some neighbourhoods are experiencing a fast residential and commercial gentrification process while others still have a derelict industrial nature with shabby housing.
The reasons why a certain NA was founded and the focus they have taken vary.Some were founded to tackle everyday problems like parking, transportation, or neighbourhood security.
Others confronted larger issues that influence an entire neighbourhood or district, such as road construction.In some cases, the motivation for founding was specific local identity (see Table 1).Today, in the situation where pressure from the real estate sector has been remarkable since the mid-2000s, all the associations have focused their activities on maintaining and improving the living environment and creating strong networks of neighbourhood activists mainly inside but also across different neighbourhoods as well as city districts, for example the Urban Idea initiative3 .
The NAs in Tallinn reflect the overall resident profile of the neighbourhood they are based in quite well.For example, the Telliskivi and Uus Maailm associations are based in inner-city neighbourhoods that are undergoing fast gentrification and the NA members present a rather good cross-section of newcomers, who tend to be younger, higher educated, and with bohemian urban lifestyles.In addition, Tallinn's inner city, with lots of green space, provides possibilities to follow a certain 'semi-rural' lifestyle (Kährik et al, 2016) while living near the city centre.In comparison a more stable resident profile-households that have lived here for generations-emerges in the Nõmme district, which has historically been a home for intellectuals and more affluent citizens.Nõmme is a garden city-style district that is characterised by large, old villa-like houses.Here there are NAs that were founded at the beginning of the 20 th century and reinstated in the 1990s, like Nõmme's Way, whose members' represent the traditional model of Nõmme, whereas Möldre Road on the outskirts of the Nõmme district unites mostly young families.An overview of all neighbourhoods and their characteristics is given in Table 2 below.
Table 2. Overview of the neighbourhoods and their residents' characteristics.
Young families, elderly, Russian-speaking long-term residents; higher education and socio-economic status, hipsters, skilled workers; private owners and tenants; Estonians.
Young families, hipsters, elderly; higher education and socio-economic status; private owners and tenants; Estonians.
Middle-aged families with higher social status, higher income, private owners; Estonians.
Middle-class to higher income families, private owners and tenants; Estonians.
Families with higher social status, e.g.university professors, long-term residents; Estonians.
Young middle-class to higher income families, house/apartment owners; Estonians and some Russian-speakers.
Young middle-class to higher income families, house/apartment owners; Estonians and some Russian-speakers.
Mixed population; young middle class families to long-term residents, Estonians and Russian-speakers, ethnic Russians; skilled workers.

Whose voice, whose interest?
According to all of the interviewees, the NAs regard themselves as representatives of local public interest, which is strongly entwined with the legitimacy issue (Plotke, 1997;Purdue, 2001;Häikiö, 2007;Connelly, 2011).The NAs deal with questions such as what their legitimacy stems from, whom they really represent, and what their inherent democracy stands for daily.We observed that NAs believe to have done much already by, for example, holding community meetings where the legitimacy questions are discussed.Consensus regarding these issues, at least among the NAs, has been obtained (Pehk and Ait, 2014).Thus, NAs in Tallinn have come to an understanding that if they do not know how the residents who are not members feel about a certain topic, they might not have the right to say that the association still represents them.Consequently, an important role the NAs have chosen to play in the local community is to be the mediators of information, which helps them to keep contacts with diverse social groups.An interviewee discusses their role as follows: The association is like a platform, an information field that allows people to bring forth different ideas.We represent this area and all residents living here purely because our activity is public, transparent, inclusive of as many residents as possible.Of course everybody cannot be included, but if they see that we are doing something they do not agree with then they can say that as well (Uus Maailm). 4verall, the voice that speaks through associations varies according to the neighbourhoods.When considering NAs separately, the force of homophily (McPherson et al, 2001;Weare et al, 2009) can be detected-in each NA certain groups prevail and the challenge to reach other groups with their activities always exists.Certain social groups-poorer and less educated-tend to be left out of the discussion (cf.Ansell and Gash, 2008) and in the case of Tallinn this is definitely an issue in inner-city gentrifying districts, where recently-arrived gentrifiers generally form the NAs.This biased representation causes neighborhoods to develop according to the visions of more active and affluent community members, as they have the capacity in terms of time and willingness to take action, suitable resources, and know-how to be included (Davoudi & Cowie, 2013).We claim that these reasons also explain why there is no apparent civil activism present in such a scope in the panel-housing estates in Tallinn.
On the other hand, when considering Tallinn as a whole, very diverse interests make their voices heard through neighbourhood activism.In some cases, geographically close communities combine their activities.For example, Pelgulinna NA activities are oriented toward the elderly and there are no clear geographical borders where NA's activities end.But regarding the issues related to neighbourhood spatial development they rely more on Telliskivi NA that actively engages with planning problems.Also, there is noticeable cooperation when organising social events such as annual neighbourhood festivities.We also discovered that interviewees see NAs as more legitimate and more influential (Häikiö, 2007) as opposed to taking action individually: 'writing as an individual did not get us anywhere, we needed a representative body.So you might say that a problem in the neighbourhood brought us together' (Liiva Village).Furthermore, they use their own individual professional experience and know-how to tackle a given issue (Häikiö, 2007), which is often influenced by more global values-using examples from western societies when arguing for their position-or seek cooperation from neighbouring NAs.
Due to the concentration of associations in certain types of neighbourhoods where the majority of the population is Estonian, the representation of Estonian and Russian voices through NAs is unbalanced.Therefore, it appears that in the dual ethnic city of Tallinn the NAs make heard primarily the voice of ethnic Estonians.This should be an alarm call for the city government as well-the neighbourhoods where community activism is limited tend to be more disadvantaged districts, lacking community resources.Lasnaidea, on the other hand, is the first neighbourhood initiative of its kind aiming to bring together residents with various ethnic backgrounds in the large minority-rich housing estate district of Lasnamäe (established as an NA in 2015).Based on the interview with the Lasnaidea representative, there are no known NAs in panel-housing estates or among Russian-speakers in general.An equivalent for NAs proposed by the interviewee might be apartment associations5 , which in high-rise panel-housing estates could have hundreds of households.Still, it is noticeable that Russian-speakers, even though not inherently less active, have not yet formed associations similar to NAs present in the inner-city areas.Furthermore, it appears that Russian-speakers do not take part in urban governance and planning discussions as actively as ethnic Estonians, meaning the voices from the Russian-speaking community are insufficient.

Connecting the disconnected
The challenge of relating people and institutions highlighted by Gaventa (2004) became apparent from the interviews.Namely, the current institutional arrangement of the Tallinn city government is seen as highly complex and internally disconnected.The interviewees stressed that often the information regarding, for example, urban planning activities is not signalled out properly, informing is just the legally required minimum.Real-life situations, however, require adjustment to particular circumstances, and sometimes special efforts should be taken to ensure that people are truly informed.The interviewees also complained that the (right) information frequently gets lost or delayed when moving between different officials.An interviewee describes the situation colourfully: The main problem of municipal offices is that the different domains and functions are so scattered that it seems different offices are like states within a state, they do not work with each other.
[…] They presume that the citizen goes from office to office with his problem and does the communication for them (Möldre Road).
Additionally, there is insufficient information available for effectively forming decisions, making decisions about what is important and requires addressing that much harder.The interviewees brought out the need to bring a more humane communication manner to the government.To build this relationship, work has to be done 'on both sides of the equation' (Gaventa, 2004, p. 26).When appropriate, the city government must explain the information it gives out in simpler and more understandable terms and the associations in turn should have adequate knowledge to ask the right questions.
In addition, it seems reasonable that as grassroots initiatives the NAs should communicate with the most appropriate level of government power.The district governments have a better overview of undergoing processes in the district, while NAs usually are more informed of the processes in parts thereof.Thus, it is important to facilitate competence of how to better engage civil associations in the district government's activities (Häikiö, 2012).If a united front comprised of local residents and the district government is created, the achieved consensus is believed to be more influential and better represented on the city level as well.
The interviewees stressed that, for NAs to be a significant partner for public bodies, a certain level of competence is essential (Innes andBooher, 2004, Häikiö, 2007;Ansell and Gash, 2008).However, the general experience of NAs regarding urban governance processes is that NAs are considered by public bodies to be annoying, a group whose opinion local government is forced to consider but whose involvement makes the process more complex and time consuming (Plotke, 1997).Thus, according to the interviewees, examples of good collaboration are a rare encounter resulting from generally insufficient informing efforts and only rare incentives from local government to involve NAs in governance processes.Interestingly, as pointed out by Ansell and Gash (2008), even though usually good previous experience encourages and bad past experience discourages collaboration, a conflict per se is not necessarily a barrier for collaboration.As the competence and trustworthiness of NAs grows through expertise as well as their self-education on germane matters, NAs consequently gain levels of legitimacy with local governments (Häikiö, 2007).Simultaneously, NAs acquire recognition as local experts or 'indicators'-the ones who provide local input.However, some sensitivity should remain on the government's part in the interest of including other civil associations that might have valuable in put in a less understandable form.By following more contextual approaches for recognition (Valentine, 2007;Fincher and Iveson, 2008), the government is more able to prevent a situation where only politically more 'suitable' associations are included in the discussion (Häikiö, 2012).

Local expertise in privatised society
According to the interviewees, the task of the city government is to balance business and citizens' interests as well as voices from different neighbourhoods.The NAs understand that the city has knowledge and vision regarding the development of the whole city.However, the interviewees argue that this vision today seems to be 'full of holes' as Tallinn's comprehensive plan is outdated and districts' comprehensive plans also tend to be vague or incomplete.This poses the question of who benefits from such vagueness.An interviewee surmises a kind of indirect conspiracy in this unclear situation: The worst problem is that the district's comprehensive plan is still not adopted.I have heard that even though the plan has not been officially approved, the Urban Planning Department has no other official document for reference, thus they act upon the [unapproved] comprehensive plan.
[…] I do not understand why this [planning] process has been stopped (Möldre Road).
In general, the NAs perceive that developers win from the unclear spatial visions of the city government, as then the 'vision' can be adjusted according to the ad hoc site-planning projects.As the majority of land in Estonia is privately owned, all planning activity is induced by someone's private interests.Therefore, the NAs feel that strong general visions in urban planning are essential.They would prefer to be involved in the initial vision formation processes, rather than in a situation in which they must oppose potentially inconsistent visions of politically influential private owners.In a 'privatised city' like Tallinn, major urban investments derive from developers.Unfortunately, neither cities nor the state have sufficient resources for large urban renewal projects in Estonia.Thus, the city has the role to set the terms of reference for the detailed plan of every new site for investment.The NAs expect to be involved in urban development projects during initial phases.It is clear that although NAs regard the local government as legitimate deciders, the NAs would prefer to see stronger accountability in the city government's actions.

Conclusions
Considering that NAs have emerged within the last decade in Estonia as neighbourhood representatives participating in urban governance, we began our research with the following assumptions concerning their goals and activities.First, by bringing forth citizen interests as local neighbourhood experts, we presumed NAs help legitimise local decisions regarding neighbourhood development, thus helping to assert citizens' political citizenship.Second, as NAs aim to bring local decisions closer to the general public, NAs' activity also opens opportunities for citizens to partake in other urban governance processes pertaining to sociourban changes, that is to assert their social citizenship.This led us to our research questions: what can we learn from the experience of Tallinn for urban governance debates in general?What are the views of the NAs on their missions, and how do they engage representation and legitimacy?
In the first phase of our research we found that the NAs did not geographically cover the entire city, meaning that only some Tallinn city districts had the opportunity to gain from neighbourhood activism.This should make the local government more cautious when making urban development decisions that could amplify the already distinctive inequalities between neighbourhoods.As NAs are inherent to gentrifying inner-city and single-family housing areas, where the socio-economic status is higher or rapidly increasing due to selective inmigration and out-migration, the local government should more systematically promote civil activism in areas where it is more demanding for citizens to instigate themselves, for example, in panel-housing estates.Unbalanced representation also causes certain neighbourhoods to develop according to the visions of more active and affluent community members, who have the capacity and know-how to be included (Davoudi and Cowie, 2013).
The NA representatives stressed that the NAs have done much work in order to create competence, and to shape themselves to be equal partners in discussions.Although expertise is grounds for legitimacy (Häikiö, 2007) it should be noted that it might lead to favouritism (Rikmann et al, 2010;Häikiö, 2012).Therefore, it is important for the public sector to provide a flexible framework for different movements-provide support for already existing forms of civil activism as well as to ensure easy access to new ones (Fincher and Iveson, 2008).In such cases, we stress that the local government should address this more systematically in order to avoid a situation where weaker voices, who might have the moral capacity but not yet the competence, i.e. they are not legitimate or 'suitable' enough (Häikiö, 2012), are restricted from discussions.Therefore, public representatives must ensure equal opportunities for NAs and individual citizens alike to secure a balanced urban governance network and to be able to recognise possible new interests in the urban sociocultural landscape (García, 2006;Valetine, 2007;Fincher and Iveson, 2008;Davoudi and Cowie, 2013).
Based on the fieldwork results, we are able to differentiate three main roles that NAs carry in urban governance in post-socialist Tallinn.These are: mediator, informant, and indicator.As NAs aim to provide a platform for fostering various ideas and activities, they are places for social encounter (Fincher and Iveson, 2008).Ideas, questions, and problems are communicated to the members of the NA, neighbourhood residents, or other people who share the ideas of the association via (social) networks that the NAs have initiated and maintained.Thus, NAs have the capacity to mediate information between the members/residents and local government.In addition, with growing competence and knowhow in urban governance-related issues, as well as acquired local expertise, the NAs are able to bear the role of indicator.By indicating certain deficiencies or imperfections in the local urban development or social issues, the aim of these activists is not to shed light on maladministration practice in the public sector or to take someone's job (as their activism is sometimes interpreted), but to provide input in constructive discussion (Häikiö, 2012).
As a post-socialist capital with a relatively young and still rather weak civil society, Tallinn shows signs of a growing movement towards becoming a city where policy decisions are made in the mode of collaborative urban governance.We might say that the marketexperiment of full private land-ownership (Tammaru et al, 2015) has evolved into a governance experiment, as the growing number of interests (namely NAs that have emerged within the last decade) regarding issues pertaining to urban development can be considered as a direct reaction to the real estate boom and fast changes in urban space.These stakeholders have opened the opportunity for public representatives to adjust their practice of governing to be more comparable to Western democracies.They aim to find ways to enrich representative practices (Plotke, 1997;Davoudi and Cowie, 2013) rather than rigidly uphold the relics of Soviet era hierarchical policies-making decisions behind closed doors within a small circle of 'experts'-by excluding public interests presented by a growing number of civil associations.
It seems that the past decade has been a time of learning about how to efficiently root collaborative urban governance practices in Tallinn.As the legislative framework was introduced more than two decades ago, both the public and third sectors have gone through extensive changes.This in turn has created a need for new forms of collaboration, which cannot be introduced overnight.We can conclude that implementing fruitful collaborative governance practices in Tallinn is still in an experimental phase as we are witnessing a generation shift in the public as well as in the third sector.Based on empirical evidence, we gather that all the necessary political characteristics of NAs-trustworthiness, accountability, legitimacy-are present or developing but the threats widely discussed in contemporary theoretical debates are noticeable also.To balance these extremes we need more experience in urban governance practices regarding how to learn better from the general public.

Figure 1 .
Figure 1.City districts of Tallinn and the location of NAs as of May 2014.Elaboration of the authors; NA locations based on Pehk and Ait, 2014.

Table 1 .
Overview of the NAs and their representatives