Anthropological studies conducted during the Russian empire categorized Estonians as Asiatic. But with the rise of nationalism, colonialism and eugenics, Estonians came to be classified – and to self-classify – as Nordic and European. Photography and painting provide a record of this visual whitening.
Getting to know the neighbours
Ethnolinguistic segregation in Tallinn districts
Estonia boasts a unified housing and labour market. So why do native Estonian and Russian-speaking residents predominantly live in separate areas of the capital? Given its current market-driven tendency, Tiit Tammaru questions whether greater cohesion or further segregation mark Tallinn’s future neighbourhoods.
Tallinn is an ethnolinguistically interesting city. Although Estonia can boast a unified housing and labour market, its capital city is segregated by language: native Estonian speakers increasingly live in different neighbourhoods to the city’s native Russian-speaking populations. Over the past three decades, an overwhelming belief in the power of the market has guided Estonia’s development. Previously, under the Soviet planned economy, even though industry played a notable role in shaping urban spatial and housing development, Estonia’s town planning was centrally organized and its populations less segregated. Now, ethnolinguistic distinctions are indicative of income disparity and residential separation. Given this starting point, two scenarios represent Tallinn’s hypothetical housing future: public intervention could facilitate greater neighbourhood cohesion or national strategies could help form neighbourhoods further separated by ethnic identity.
Breaking down ‘segregation’
Residential segregation or differentiation indicates the amount or rather lack of social cohesion and integration in a city. Often associated with social problems, segregation is considered dependent on three important preconditions that can restore greater housing fairness: first, the residential preferences of two groups with different mother tongues must be similar; second, the groups must have equal income or purchasing power; and third, there should be no discrimination based on ethno-cultural characteristics.
Within an academic context, however, segregation is a neutral term. It indicates the extent of differences in the distribution of population groups across urban neighbourhoods. In this case, the differences in question gauge the distribution of Estonian versus Russian speakers across Tallinn. A dissimilarity index ranging from zero to 100 is frequently used to measure the spatial distribution of groups across urban neighbourhoods, where zero would denote the equal distribution of linguistically different residents and 100 that each neighbourhood is inhabited by representatives of only one language group. As a continuous indicator, values actually come nowhere near these extremes. They tend to range from 20 to 70 and rarely surpass 50. High index values are characteristic of urban centres with strong demarcation lines between different groups such as large US and Latin American cities. And the numerical interpretation of the index is simple: it shows the share of members from one or another group who should move to a different neighbourhood in order to achieve the same distribution of groups across the city.
Tallinn is a particularly interesting case study because its inhabitants are divided into two almost equal groups: the population approximately constitutes half native Estonian speakers and half native Russian speakers. If there was no residential segregation, the city’s neighbourhoods would be divided 50/50 between the two linguistic groups. Theoretically, each apartment could be equally accessible to native Estonians and native Russian speakers. And each work environment, each school, each recreational centre could be perfectly mixed with a zero dissimilarity index.
When opposites don’t attract
While the dissimilarity index may be easy to interpret numerically, its substantive interpretation is much more complicated. Usually, the greater a city’s segregation, the greater its social problems are perceived to be. However, such a clear cut interpretation can be overly simple. One of the mechanisms that affect residential differences is homophily (i.e., similar people attract one another). Indeed, the mutual attraction of those who are alike is a very common behaviour pattern in social relations because it is often easier to get along with people with a common language or culture, similar levels of income, similar interests or other common features.
Imagine the following scenario: a person with cultural interests enters a room occupied by two people, one interested in art, the other in sport. The people with cultural interests have a common interest that links them and makes them more likely to interact. This is what the Nobel award-winning American economist Thomas Schelling has generalized into a dynamic model of segregation: even a slight individual preference for people to interact with those similar to themselves will lead to very high levels of segregation between population groups. Like attracting like may seem the perfect complement, but it also creates distance: in our example, when the people with cultural interests start interacting, they leave the sporty person alone; the dark side of the homophily principle can lead to the distinctive person being unconsciously pushed out, potentially revealing elements of discrimination.
Most of Tallinn’s Estonian speakers ethnically identify themselves as Estonians, whereas the city’s Russian speakers display more diverse ethnic identity, with ethnic Russians forming around 8 out of 10 people in this group.
Tallinn’s all-mighty market
As in other eastern European countries, Estonia’s housing pendulum swung from centralized urban planning to market-led development in the 1990s. Ideological considerations underpinned the desire to make a clear break from the mentality of ‘what belongs to everybody belongs to nobody’, inherited from the planned-economy past, and to underscore a sense of responsible ownership. Pragmatic considerations, on the other hand, were related to an impoverished public sector. In a situation where there were no funds even for refuelling police cars, it was easier to hand the public sector’s greatest item of expenditure – housing – over to the people.
Eastern European countries rank among the first in the world for their proportion of apartment owners and Estonia’s ‘super homeownership society’ is no exception. Even in liberal market economies such as the US, the proportion of apartment owners is markedly lower than in Estonia. The country’s approach to apartment management also promotes owners in a central role: owners and the associations they form generally manage entire apartment blocks. Renovation of the existing pool of property – known as ‘DIY urbanism’ – is widespread. Visually, this is expressed most clearly in the façades of houses where diverse windows and balconies reveal the different tastes of individual apartment owners. New edifices, however, are built according to the wishes of developers, a practice known as ‘investor urbanism’.
The coexistence of an uneven renovation of the existing pool of properties and investor urbanism has caused great variations in the quality of living spaces on offer. Although fault has been found with the construction quality of new buildings, both new and completely renovated small blocks enjoy a markedly higher repute among buyers than Soviet-era apartments. And those apartments built at the height of socialist modernism in the 1970s and 1980s are held in the lowest esteem.
In the 1990s Estonia’s income gap grew. And native Estonian speakers have been able to make faster progress than native Russian speakers in the Tallinn labour market. Money buys choice on the housing market and, therefore, increased income differences between Estonian and Russian speakers is one important factor behind greater residential segregation. Estonian speakers are over-represented among the inhabitants of new and renovated houses, whereas Russian speakers are over-represented among the inhabitants of houses built in the 1970s and 1980s. As a result, the dissimilarity index value summarizing residential segregation between Estonian and Russian speakers has grown from 31 in 1989 to 44 in 2019. In other words, almost 1/3 of Estonian or Russian speakers would have had to move home in order for the distribution of both groups to be similar in the various Tallinn neighbourhoods in 1989; in 2019, however, almost half of Estonian or Russian speakers in Tallinn would have had to do the same to achieve a similar result.
Changes in Tallinn’s ethnolinguistic landscape become even clearer if we look at city districts. In 1989 the proportion of Estonian and Russian speakers was close to the entire city’s average in most districts, fluctuating between 40% and 60%. The only exception was constituted by the two least populated districts, Pirita and Nõmme, which had a large majority of Estonian speakers. Today, however, the situation has changed substantially. Districts with a clear majority of native Estonian speakers and others with a clear majority of native Russian speakers have emerged in Tallinn. In some instances, districts themselves can be further grouped into larger areas according to the changes they have undergone (i.e., changes are taking place not only at neighbourhood and district levels but also on a higher spatial scale).
Over the past thirty years, Russian speakers have converged on Lasnamäe. Proportionately speaking, a third of native Russian speaking Tallinnites lived in Lasnamäe in 1989, whereas almost a half did in 2019. At the same time, the proportion of Russian speakers as a percentage of the whole population of Lasnamäe grew from 64% to 75%. And changes have occurred within Lasnamäe itself: the proportion of native Russian speakers has most notably grown in areas of the district located furthest away from the city centre.
The other city district where the proportion of native Russian speakers has grown even more rapidly is Pirita. In Soviet times, every tenth inhabitant of Pirita spoke Russian as their mother tongue. Today, every fourth resident does. Even among native Russian speakers themselves, segregation is happening according to wealth: more affluent families move away from Lasnamäe, commonly into a private house, yet remain in the same area of the city within Pirita or move to Muug, a garden suburb right outside the city’s limits. In broad spatial terms, the proportion of native Russian speakers in Tallinn’s eastern districts has grown especially quickly.
The precisely opposite trend is characteristic of Tallinn’s city centre or its Central, North Tallinn, and Kristiine districts. At the end of the Soviet era, these three districts had more native Russian speakers (38%) than Lasnamäe (33%); today, however, the proportion of native Russian speakers has dropped to one fourth of the population. Only the tip of the Kopli peninsula still houses a majority native Russian-speaking community; and yet, sooner or later, the large housing developments currently underway there will begin to change the ethnic and linguistic composition of that district’s population too. In other words, the proportion of Estonian speakers is growing in the city centre and the urban regions bordering on the Central district, both in the directions of North Tallinn and Kristiine and, to a lesser degree, also Lasnamäe.
A third, large, notional urban area is constituted by West Tallinn, where, on the whole, the proportions of Estonian and Russian speakers have not changed very much over the past thirty years. However, interesting shifts have been happening in this part of the city too. The proportion of Russian speakers is greatest at Astangu, constituting upwards of 70%. Likewise, the proportion of native Russian speakers has grown to more than 50% in Haabersti. Even greater segregation can be observed in the areas dominated by private houses. As in Pirita, the proportion of Russian speakers has grown in Kakumäe, whereas, in Nõmme, on the contrary, the proportion of Estonian speakers has risen and only few Russian speakers live there.
The market’s invisible hand has distributed native Estonian and Russian speakers into different areas of Tallinn both on neighbourhood, district and pan-district scales. However, the future is unlikely to cause noteworthy growth in the segregation index. As noted above, the numeric value of the index rarely rises above 50. The concentration of Russian speakers at Lasnamäe, which houses every fourth Tallinite, is already extensive, and, in order for levels of residential segregation to keep growing, Estonian speakers would have to abandon that part of the town altogether. This, however, is very unlikely, as the district itself is not homogeneous: its areas between the city centre, Kadriorg and urbanized Ülemiste, including the newly revived neighbourhood of Pae park, are highly attractive residential areas. Wealthier native Russian speakers, in their turn, seek exclusive housing in districts with private houses such as Pirita and Kakumäe, where many Estonians live. The only large district where the majority of native Russian speakers still has potential to grow is Väike-Õismäe, but this area already has far fewer inhabitants than Lasnamäe and, therefore, the impact of changes would not be comparable.
A coherent society
One alternative and potential future scenario would involve public intervention to reduce segregation. Estonia’s central governmental policy currently emphasizes ethnic integration and social cohesion but does not directly exert influence on Tallinn’s neighbourhood configuration. A social cohesion scenario would aim to achieve a more even distribution of Estonian and Russian speaking residents throughout the capital.
So far, the government’s integration measures have concentrated on improving language skills. This is understandable. As a national language, Estonian is linked to identity and administrational ease. However, an integrated and coherent society requires more than knowledge of the official language. It also involves cultural knowledge, shaping a common identity through interaction, including learning, working and spending leisure time together, living in the same areas, interethnic cohabitation, and forming friend groups and communication networks. People who learn Estonian while living in segregated communities do not develop common social networks nor a common information space. Learning a language in isolation is not as fruitful as joining in active communication with native speakers. In short, an activity-based approach that brings members of different ethnolinguistic groups together is more effective. This has been understood at a state policy making level, at least in relation to language learning.
Segregation according to native languages is caught up in a vicious circle tying residential neighbourhoods, workplaces and schools together. The circle begins at birth. A child’s parental home determines the neighbourhood where they will grow up, the friends they will meet and school they will attend. Estonia’s parallel education system adds an additional factor: language segregation begins at kindergarten, where children are divided into Estonian and Russian language schools. In other words, children are sorted into different educational paths based on their mother tongue. Although not all Russian children go to Russian schools and the level of Estonian studies in Russian-language schools has improved, many children still study in separate schools and do not cross the language border.
Job prospects are determined not only by schools and education but also by acquaintances and friends; many positions are still obtained through word of mouth even in relatively equal societies such as Sweden. In Estonia, the differences in labour market outcomes between Estonian and Russian speakers are strongly linked to the ethnic and linguistic distribution within Estonia’s education system.
Labour market success, in turn, influences housing market prospects. When a couple buys a home together, their budget is jointly determined. As mixed marriages are rare in Estonia, the average Russian-speaking family’s housing budget is much lower than that of the average Estonian-speaking family. Those couples who have children provide a parental home that influences their educational opportunities and the vicious circle starts again.
The framework of this housing trap may be a broad generalization – real life, after all, is more diverse – but it does help to explain the broad picture. In addition, certain other factors are known to magnify and reinforce existing patterns. For example, parents may choose their home location according to the school where they want their children to study. This may bring Russian-speaking families closer to Russian-language kindergartens and schools, in turn motivating Estonian-speaking families to move away. These mechanisms have not been studied in Estonia, but research from other countries such as neighbouring Finland and Sweden shows that the main ethnic group’s residential choices tend to reinforce segregation. More specifically, these studies show such behaviour often causes levels of residential segregation to grow in ethnolinguistically diverse cities. And this strategy is most common among families with children.
As a first step, the coherent society development option assumes a levelling off of income differences measured by the Gini index. The Estonian state’s modification of taxation policy has recently helped reduce income differences, which has had a positive impact on residential segregation levels. Additional steps that could be taken within the labour market would range from equal levels of employment to reducing the differences between fields of activity available to employees according to their mother tongue.
However, if low-level residential segregation is an indicator of integration and the labour market is its executor, then school is its launchpad. As long as children are distributed to different schools according to their mother tongue, differences between ethnic groups based on labour and housing markets will remain in place. A key issue to be resolved here is how to organise Estonia’s education system in a manner that would contribute to a more coherent and integrated society, thereby helping to reduce differences in social networks, and labour and housing markets.
Leisure activities also influence integration, depending on whether they are predominantly based on language or common interests and where they take place. Some leisure activities lead to high levels of segregation. A well-known example recognizes that cities are more culturally and spatially divided on Sunday mornings when Christian residents go to church. In contrast, many leisure activities such as sports, sporting events or large music concerts bring people together regardless of their mother tongue.
Over the past thirty years, the leisure activities of Estonian and Russian speakers have on the whole become more similar, but the places where they practise them are still different. This factor could be influenced on a municipal level. The arsenal of possible interventions is rich. At one extreme, large cultural venues can be strategically placed in certain neighbourhoods to encourage people from all over the city to prescribed locations. Not all important venues need to be located in the city centre. A new Tallinn opera house, for instance, could be built in Lasnamäe. This would bring new activity, people, attention to the district, improving its reputation. And an altered reputation is often key for change. At the other extreme, small everyday activities such as learning groups can bring native Estonian and Russian-speaking children together. Bringing children together helps bring parents together and the snowball of interethnic communication quietly starts rolling. As interethnic social interaction grows, so does the knowledge of Estonian.
Cities have been culturally, linguistically and ethnically diverse throughout history. Often different groups inhabit certain districts: Chinatown is probably the most well-known example of a distinct ethnic neighbourhood in many cities. Another potential scenario, which is the polar opposite of a strategy for cohesion, accepts cultural difference and existing segregation. The ‘diversity enriches’ scenario, based on homophily or like attracting like, promotes culturally distinct urban neighbourhoods. The tools for this approach are found at a national rather than municipal level.
In the US, the term ‘ethnoburb’ – a combination of the words ‘ethnic’ and ‘suburb’ – has been coined to denote linguistically and culturally diverse suburbs where one ethnic group predominates numerically over others. In Los Angeles, where the term was first adopted, it has been primarily used to signify the shift of Chinese immigrants from the centre of the city to the suburbs. Ethnoburb infers that linguistic and cultural attraction play a more important role than income in the formation of ethnolinguistically homogenous neighbourhoods. Ethnoburbs are not socially problematic residential areas. However, they may become problematic as evidenced, for example, by the suburbs of French and Swedish cities.
Tallinn’s spatial developments have led to ethnically more homogenous neighbourhoods with no major social problems, at least so far. Lasnamäe, especially at its more distant edges, has attracted a population strongly dominated by Russian speakers over the last thirty years. Certainly, wealthier Russian speaking families have moved away from Lasnamäe, for example to Pirita, but the broader trend has still been their convergence on Lasnamäe. And, since the Russian speakers’ community is ethnically and culturally diverse, the district is a rich mix reflected in the diverse cuisine on offer, which is nevertheless a hidden attribute not known city-wide.
Other important features linked with Russian language and culture have converged on the outer edge of Lasnamäe, one of the most important being the newly enlarged and completely renovated Tallinn Linnamäe Russian Lyceum. The ‘Quick to Hearken’ Icon of the Mother of God church, completed in 2013, together with Patriarch Alexius II Square, is another positive development. Both buildings are situated a couple of kilometres apart along the natural coastline’s cliffs.
Russian high schools, currently perceived as a division within Estonia’s educational system, could be reconsidered as a positive cultural cornerstone. Tallinn has a Lycée Français and an English College, both of which are very highly regarded by local children irrespective of their ethnic background. Russian high schools have the potential to evolve into highly esteemed centres of Russian culture where people from all family backgrounds might want their children to study.
Conceptualizing segregation as an opportunity rather than a problem places issues of neighbourhood reputation at the centre of urban discussions. City districts from the modernist era, a.k.a. the ‘mountains’ of Tallinn, are most evidently challenged by the social downgrading that has occurred across Europe. Thirty years ago, the district of North Tallinn had a bad reputation. It was known as ‘The Rot District’, because it was inhabited by the capitals poorest people. Nowadays, in contrast, it is one of the most highly esteemed residential areas of the entire city. Several beneficial approaches associated with modern city districts – such as the green turn, sustainable living, reducing the ecological footprint and car use, compact cities and the sharing economy – could be aptly applied to Tallinn’s ‘mountains’. Public space may have generally been improved throughout Tallinn, but district apartment blocks, as a rule, have not been overhauled. Rather than relying on DIY-style solutions, city renovation could also be carried out on a more systematic scale. The Estonian Art University’s Unfinished City project offers one such vision of what modern urban space at Lasnamäe might look like. Scientists at Tallinn Technical University have shown that, thanks to standard house types, buildings can be renovated so that a large part of the preparatory work is done in factories (i.e., prefabrication techniques are applied to renovations). This would entail a marked reduction of the time spent on building sites, making renovations both cheaper and less nerve-racking for inhabitants.
Whether Tallinn’s neighbourhoods continue along their market-driven trajectory, aim for more social cohesion or further embrace cultural division is open to debate. The best potential outcome requires public consultation on these options. The city’s native Russian and Estonian speakers have largely moved into different districts over the last thirty years. Only public debate can establish if this is a problem or an opportunity. The major question that local inhabitants, communities and their elected leaders need consider is to what extent Tallinn’s development should be led by the market or could be envisioned by the people that it serves.
 For example, the labour market and labour market outcomes such as salaries must not be differentiated according to the mother tongue of workers.
 T. C. Schelling, ‘Dynamic Models of Segregation’, The Journal of Mathematical Sociology, 1971, Vol. 1, No. 2, pp. 143–186.
 M. Stephens, M. Lux and P. Sunega, ‘Post-Socialist Housing Systems in Europe: Housing Welfare Regimes by Default?’, Housing Studies, 2015, Vol. 30, No. 8, pp. 1210–1234; J. Salukvadze and O. Golubchikov, ‘City as a Geopolitics: Tbilisi, Georgia – A Globalizing Metropolis in a Turbulent Region’, Cities, 2016, Vol. 52, pp. 39–54; D. Kiss, Modeling Post-Socialist Urbanization: The Case of Budapest, Basel, 2019.
 All data used in the article come from Infotechological Observatory of Mobility (Infotehnoloogiline Mobiilsusobservatoorium, https://imo.ut.ee/). Kadi Kalm, research fellow at Tartu University, helped me with the calculations. The research was supported by grant PRG306 of the Estonian Research Council.
 Russian schools provide language immersion classes and, at secondary school level, a number of subjects are taught in Estonian.
 A. Forslund, Employment Outcomes and Policies in Sweden During Recent Decades (IFAU Working Paper 2019:15). Uppsala, 2019.
 K. Leetmaa, T. Tammaru and D. B. Hess, ‘Preferences Toward Neighbor Ethnicity and Affluence: Evidence from an Inherited Dual Ethnic Context in Post-Soviet Tartu, Estonia’, Annals of the Association of American Geographers, 2015, Vol. 105, No. 1, pp. 162–182.
 C. Peach, ‘Social Geography: New Religions and Ethnoburbs – Contrasts with Cultural Geography’, Progress in Human Geography, 2002, Vol. 26, No. 2, pp. 252–260.
 https://www.ttu.ee/ttu-ehitab-vana-uhiselamu-liginullenergiahooneks; https://www.postimees.ee/4079085/mustamael-katsetatakse-paneelmajade-paasteplaani.
Published 12 October 2020
Original in Estonian
First published by Vikerkaar, June 2020
Contributed by Vikerkaar © Tiit Tammaru / Vikerkaar / EurozinePDF/PRINT
Estonian journal Vikerkaar on gendered domesticity and the double standards of state socialist gender politics; neo-Ottoman modernization and luxury development on the Belgrade waterfront; post-Soviet housing projects in Tartu; the philosophy of single-family homes, and more.