Research results on segregation in the Helsinki metropolitan area indicate quite significant differences between sub-city areas in terms of income and education levels, percentages of immigrant population and employment rates (e.g. Vaattovaara & Kortteinen 2012; Vilkama 2012; Lönnqvist & Tuominen 2013). Although the general increase in education and income levels is evident across almost all areas, the differences have remained the same – or even increased – over the past decade (Vilkama et al. 2014). The increase in area differences has always been seen as a negative phenomenon, and various political measures have been taken to curb or reverse it. On the other hand, it would seem that social segregation is a fairly common and often permanent urban phenomenon.
The housing stock and housing market play a central role in social segregation (Cheshire 2006). The stock of social rental housing, the proportion of which in the entire housing stock varies between areas, is often categorically reserved for those on lower incomes. As regards the housing stock priced on market terms, the type and location of the buildings, when capitalised in the prices and rents of the housing units, have a significant effect on the socioeconomic structure of residents. The varying solutions employed by municipalities in planning and land use policy and housing policy may also lead to varying results in resident structure. Furthermore, different demographic groups may have their own wishes in terms of the socioeconomic structure of the residential area. The above-mentioned characteristics of the housing stock and housing market, together with the income and wealth of households, form the limitations within which these choices are made.
In this article we will examine the development of the socioeconomic status of small areas in the Helsinki metropolitan area. The intention is to find out how permanent their social status is. After that, we will look into the possible connection of the structure of the housing stock and the location of the area to the socioeconomic status of an area. We will seek to determine the degree to which the structural characteristics and location of the housing stock – which change slowly or not at all – explain the socioeconomic status of an area. The article is based on a more extensive research report on area development in the Helsinki metropolitan area, which was published in the spring of 2014 (Vilkama et al. 2014).
Incorporating the dimensions of social segregation into one index
The growth of spatial differences is often approached through individual dimensions of segregation, one variable at a time, as it were. Analyses of individual dimensions produce interesting and topical information on the development of residential areas, but it is difficult to form an overall view of the wider developments in the areas, or changes in their status, based on them. In this article, we will utilise the area status classification developed by the City of Berlin, which condenses the dimensions of spatial segregation into a single variable depicting area status.
Since 1998, the City of Berlin has utilised and developed various indices for monitoring social development in residential areas. The purpose of the indices has been to provide tools to support the monitoring of area development, and the focusing and planning of various regional measures. At the moment, three separate indices are in use: the status index describes the status of areas as a summation of various indicators; the development index depicts the perceptible dynamic of area development; and the social development index incorporates the values of the two other indices into a single variable (see Social Urban Development Monitoring 2010
Within the scope of this article, we will apply the first of the indices employed by the City of Berlin – the status index – which condenses the dimensions of spatial segregation into a single variable and enables us to view the status changes of areas between 2002 and 2012. In this context, status changes refer to changes in the order of the areas; in other words, changes in the status of an area in relation to other residential areas.
The status index was calculated on the basis of four variables measuring socioeconomic and ethnic segregation in residential areas: the proportion of 25- to 64-year-olds with only basic education; the unemployment rate of 25- to 64-year-olds; earnings subject to national taxation per each resident aged 15 or over (i.e. average income in the area); and the share of the population with a foreign mother tongue. The variables we used deviate from the statistical variables used in the City of Berlin index, but the calculation method is similar.
The areas were arranged in ascending order, after which the variable-specific ranking points (the ordinal ranking numbers of the areas for each variable) were added together. Thus the areas that ranked the lowest on several variables received the most points. The areas were then classified into ten groups of equal size. The areas in the lowest decile were named as ‘low status’ areas, in line with the Berlin classification; the areas in the second lowest decile were dubbed ‘moderately low status areas’; a ‘middle’ class was formed from the middle 60 per cent of the areas, and the final 20 per cent were placed in the ‘high status’ class.
The classification was conducted according to the situation at the turn of 2011/2012 and 2001/2002, so as to be able to examine the changes in status from 2002 to 2012. The area level used in the analysis was the smallest possible statistical area where data were available: sub-districts (osa-alue) in Helsinki, and comparable statistical areas in Espoo (pienalue) and Vantaa (kaupunginosa). New residential areas constructed in the 2000s, or areas that had fewer than 300 inhabitants during any part of the ten-year period, were excluded from the categorisation. Likewise, the parts of Sipoo that were annexed into Helsinki in 2009 were not included since statistical data was not available prior to the annexation. Thus we had the same number of areas throughout the ten-year period – any changes in the number of areas would have also affected their placement into the status classes. In total, the study encompassed 226 areas in the Helsinki metropolitan area.
Area status appears fairly constant
The status index provides a picture of the social status of residential areas in the Helsinki metropolitan area which is very similar to the results of previous studies on segregation processes in the region (e.g. Maury 1997; Vaattovaara 1998; Kortteinen & Vaattovaara 1999; Kortteinen et al. 2005; Vilkama 2012). Areas with weaker status (the lowest fifth; low and moderately low status) are primarily dominated by blocks of flats and situated along railway tracks and metro lines. High status areas (the strongest fifth), in turn, are located by the sea, in the eastern parts of Espoo and in some areas dominated by single-family houses in various parts of the metropolitan area. For example, the majority of southern and western Helsinki falls into the middle status class.
Furthermore, the differences in social status of small areas between the three cities are evident in the manner indicated above (see e.g. Lönnqvist & Tuominen 2013). Espoo stands out from Helsinki and Vantaa as a municipality of significantly higher status (Table 1). Only four of the residential areas in Espoo are in the two weakest status classes, which include a total of 45 areas in the region (20 per cent of all areas). The residential areas in Helsinki are more evenly distributed among both extremes of the status classification. However, areas of Helsinki are clearly overrepresented in the lowest status class. At the same time, Helsinki includes a number of high status areas. The districts of Vantaa, on the other hand, typically appear as middle- or moderately low-class areas. Only three of the Vantaa districts are in the high status class.
As can be seen, the spatial structure and inter-municipal differences depicted by the status index are quite clear. But how permanent do the area statuses appear if the changes in status are viewed over a span of ten years?
Table 2 describes the status classes of the residential areas in the Helsinki metropolitan area at the turn of 2001/2002 and 2011/2012. The majority of the areas (83 per cent) remained in the same class, and only less than one-fifth (17 per cent) moved from one class to another. For example, almost all (19 out of 22) of the areas in the low status class at the turn of 2001/2002 are still in the same class a decade later. This is the case despite the fact that the general increase in education and income levels has also had a positive impact on the development of these areas. However, in three low status areas, development has been more positive than average, leading to these areas (one from each city) climbing to a higher status class. As a result of this, the status of three other areas has weakened correspondingly, and they have fallen to a lower status class (two areas in Helsinki and one in Espoo). Similar minor transitions from one class to another also occurred in the high and middle class. Overall, the status of all areas appears to have remained highly constant. In summation, we can therefore state that the weakest areas have remained the weakest and the strongest areas have remained the strongest.
Impact of housing stock structure and area location on socioeconomic status
In the following, we will examine the impact of the housing stock structure and the location of residential areas on their socioeconomic status. Above, socioeconomic status was described using a status index (‘Berlin classification’) based on the ranking of areas. In this section, however, we will apply an approach based on main component analysis, since in the ranking-based examination of the data, differences between areas are only linked to the placement (ordinal ranking number) of the residential areas on the ranking list. The difference in status between the areas that rank 10th and 20th is equal to that between the 30th and 40th. As regards the variables behind the ranking order, the differences are not equal in this way. If we look, for example, at the proportion of foreign language speakers, the differences between areas in the middle status group are fairly minor, but the differences between the areas in the group with the highest proportion of foreign language speakers are naturally larger. Main component analysis allows us to take into account these types of variations in the distribution of basic variables.
Comparing the result produced by the status index to the results of the main component analysis conducted with the original variables indicates that the ‘Berlin’ classification provides a fairly good description of the sub-city level variations in socioeconomic standing. The correlations between the original variables – proportion of 25–64-year-olds with only basic education; unemployment rate of 25–64-year-olds; earnings subject to national taxation per person (aged 15 or over); and the share of foreign language speakers – are relatively strong and highly significant in statistical terms (Table 3). For example, there is a negative correlation between the proportion of foreign language speakers and the average income in the area, which stands at 0.47. The correlations between other variables are even stronger. These strong correlations indicate that the variables produce a very similar regional picture where good and poor social standing are located in largely the same areas.
The area differences depicted by the four basic variables can, through main component analysis, be condensed into a single dimension that explains 72 per cent of the variation of the variables between the areas. If the result of the main component analysis – a score describing the socioeconomic status of the areas – is viewed on the map together with the previously presented status classification (‘Berlin classification’), it can be seen that the spatial picture is very similar with both methods.
At the outset of this article, we posed the question of how the housing stock structure and location of residential areas affect their socioeconomic status. We base our examination on a regression model in which the socioeconomic status is explained through the housing stock structure (proportion of ARA dwellings, proportion of flats) and location of the area (distance from the centre of Helsinki; municipality.)
Based on the estimation results of the regression model (Table 4), the above-mentioned variables describing the structure and location of an area explain approximately 68 per cent of the variation in the socioeconomic status of residential areas. The share of flats and ARA housing units alone explains more than half (51 per cent) of the variation between areas. According to the results (Table 4), the shares of ARA housing and flats both lower the status of an area. The latter may seem somewhat peculiar as the central area of Helsinki, which is the most expensive area in the region measured in price per square metre, is composed mostly of blocks of flats. The result makes more sense when the distance from the city centre is included in the equation. An increase in the distance dramatically lowers the socioeconomic status of an area. Therefore, among two residential areas at the same distance from the city centre, the one with the higher share of single-family houses will have the higher socioeconomic status, on the basis of the model applied here. In addition to this, there are significant differences between municipalities. Being located in Espoo will substantially increase the prognosis of an area's socioeconomic status according to the model. This is in comparison to Helsinki and Vantaa, as a statistically significant difference cannot be seen between the latter two.
What this means is that distance from the city centre, housing stock structure and municipality can be used to account for a significant portion of the variation in social status between areas. Therefore, it is natural to conclude that the permanence of social status, which was also examined above, is largely explained by these structural factors. After all, changes in housing stock take a long time. For example, at the city level, newly produced housing typically amounts to 1 per cent of the total housing stock.
In this article, we tested the possibilities provided by the status index developed by the City of Berlin for analysing the development of regional differences in Helsinki. Based on the results and a comparison with the results of the main component analysis utilised towards the end of the article, it can be seen that the classification provides a feasible way to illustrate socioeconomic segregation in the Helsinki metropolitan area. The results are also in line with those of previous studies (such as Maury 1997; Vaattovaara 1998; Kortteinen & Vaattovaara 1999; Vilkama 2012; Lönnqvist & Tuominen 2013). At least within the ten-year period investigated in this article, area status seems fairly constant. Changes are minor and primarily occur in relation to the basic variables behind the index between areas that are similar to each other. Here, however, lies a point of criticism regarding this method. How authentic are status changes when the rankings of very similar areas can be changed as a result of minor alterations in the basic variables?
When the housing stock structure and location of an area is used to explain its socioeconomic status, we gain very clear and possibly expected results. The fact that a high proportion of ARA housing lowers socioeconomic status is a given, since the income criteria utilised in the selection of tenants for this form of housing steer the development in this direction. The role of the distance from the city centre is also clearly evident. The status of an area is lowered (building type and municipality standardised) with increasing distance from the city centre. The differentiation of Espoo (other factors standardised) from Helsinki and Vantaa is also clear. Due to the slow change of housing stock, the socioeconomic structure can also be expected to change slowly.
On the other hand, as valuations of housing change and the urban structure develops, the valuation of individual areas may also change. This may occur as a result of the strong population increase of the city and related large-scale construction of housing in both new and old areas. Although the status of areas appears cast in concrete in the short-term – even in the span of ten years – area status may change significantly over a longer period of time. Who, for instance, would have thought 50 years ago that Käpylä or Punavuori would emerge as desirable residential areas?
Katja Vilkama is Senior Researcher at City of Helsinki Urban Facts. Henrik Lönnqvist is Researcher at City of Helsinki Urban Facts.
Cheshire, Paul (2006). Resurgent Cities, Urban Myths and Policy Hubris: What We Need to Know. Urban Studies 43: 8, 1231–1246
Kortteinen, Matti & Vaattovaara, Mari (1999). Huono-osaisuus pääkaupunkiseudulla 1980- ja 1990-luvuilla – käännekohta kaupunkiseudun kehityksessä? Terra 111: 3, 133–145
Kortteinen, Matti, Vaattovaara, Mari & Alasuutari, Pertti (2005). Eliitin eriytymisestä pääkaupunkiseudulla. Yhteiskuntapolitiikka 70: 5, 475–487
Lönnqvist, Henrik & Tuominen, Martti (2013). Asuinalueiden sosiaalinen kehitys. In: Helsingin tila ja kehitys 2013, 171–175. Helsingin kaupungin tietokeskus, Helsinki
Maury, Jon (1997). Ilmakuva pääkaupunkiseudun asuinalueiden sosiaalisista eroista 1990-luvun alussa. In: Ilmonen, Mervi (ed.): Mitä osoite osoittaa? Asuinalueiden erilaistuminen Helsingin seudulla, 55–88. Pääkaupunkiseudun julkaisusarja B 1997: 2
Social Urban Development Monitoring 2010. http://www.google.fi/url?sa=t&rct=j&q=&esrc=s&frm=1&source=web&cd=1&ved=... (viitattu 5.12.2013)
Vaattovaara, Mari (1998). Pääkaupunkiseudun sosiaalinen erilaistuminen. Ympäristö ja alueellisuus. Tutkimuksia 1998: 7. Helsingin kaupungin tietokeskus, Helsinki
Vaattovaara, Mari & Kortteinen, Matti (2012). Segregaatiosta ja sen inhimillisestä ja yhteiskunnallisesta merkityksestä. Talous & yhteiskunta 2012: 3, 60–66
Vilkama, Katja (2012). Socio-economic and ethnic differentiation of neighbourhoods in Helsinki. Helsinki Quarterly 4/2012, 24–31
Vilkama, Katja & Lönnqvist, Henrik & Väliniemi-Laurson, Jenni & Tuominen, Martti. Erilaistuva pääkaupunkiseutu. Sosioekonomiset erot alueittain 2002–2012. Tutkimuksia 2014: 1