Neighbourhood-based social integration . The importance of the local context for different forms of resource transfer

*Corresponding author: Prof. Dr. Andreas Farwick, Ruhr-Universität Bochum, Geographisches Institut, Universitätsstraße 104, 44799 Bochum, Deutschland, E-mail: andreas.farwick@rub.de Heike Hanhörster, Isabel Ramos Lobato, ILS – Institut für Landesund Stadtentwicklungsforschung, Brüderweg 22-24, 44135 Dortmund, Deutschland Wiebke Striemer, Ruhr-Universität Bochum, Geographisches Institut, Universitätsstraße 104, 44799 Bochum, Deutschland Abstract: Due to their lack of financial resources, poor residents of deprived neighbourhoods are very much reliant on support and assistance from their personal networks. Studies refer to the key importance of neighbourhood contacts transcending social boundaries to promote upward social mobility. Based on a mix of quantitative and qualitative findings, this paper looks at the importance of social mix within a person’s neighbourhood and immediate surroundings for transferring different kinds of resources. The results show that even residents of deprived neighbourhoods can call on a well-developed support network to deal with everyday problems. The contribution also shows that network contacts to people endowed with more resources are no guarantee for the upward social mobility of the less well endowed. Indeed, it would seem that ‘getting-ahead’ resources are also accessible via their homogeneous networks. Much more to the point, the immediate surroundings turn out to be an important spatial context for contacts and resource transfers, especially for families with children.


Introduction
Increasing social and ethnic diversity is leading to changes in the conditions under which people live together in German cities. As in many other European countries, we have long been observing increased social-spatial polarisation and an associated increase in social problems (Friedrichs/Triemer 2008;Farwick 2012;Schönwälder/Petermann/Hüttermann et al. 2016; Musterd/Marcińczak/ van Ham et al. 2017). One recent study provided data on the rise in social segregation in German cities since 2002, highlighting the fact that it is increasingly seldom to find different social groups living next door to each other (Helbig/Jähnen 2018: I). The question thus arises as to what effects this increasing segregation has on neighbourhood life and on residents' access to resources.
The increasing concentration of people experiencing poverty in specific urban neighbourhoods has led to discussion of the socially disadvantageous effect of these neighbourhoods on their residents (Friedrichs 1998, Farwick 2001; van Ham/Manley/Bailey et al. 2012;Galster/Sharkey 2017). As research illustrates, negative neighbourhood effects can for instance be passed on to residents on account of a lack of resources, the dominance of negative role models, stigmatisation and discrimination (Farwick 2012: 391 ff.;cf. Galster 2012).
At the same time, however, this local neighbourhood level offers low-income households in particular an important reference space influencing their social integration (in the sense of access to functional, social and symbolic resources) through interaction with other residents (cf. Forrest/Kearns 2001). In this context, special importance is accorded to resource transfers transcending social boundaries (i.e. between people of different social status). Several scholars argue that such transfers have the potential to greatly reduce the extent of social disadvantage (Putnam 2000;Pinkster 2009;Pinkster/Völker 2009).
As a way of promoting resource transfers between different social groups, Germany and other (European) countries tend to follow strategies based on socio-spatial mixing, assuming that the spatial proximity of people of different social status also leads to social proximity, with poor residents benefiting from contacts to higher-income residents through the transfer of resources. However, empirical findings regarding this assumption are not as yet conclusive (cf. Bridge/Butler/Lees 2012; Weck/ Hanhörster 2015). Moreover, there is no firm evidence regarding the spatial level at which mixing promotes the development of contacts transcending social boundaries. Studies undertaken by Hanhörster (2015: 3114 f.) and Farwick (2009: 230 ff.) suggest that it is not so much the social composition of the neighbourhood but that of the immediate surroundings that determines the extent of boundary-crossing contacts. Opportunity structures for meeting up with other people in one's immediate surroundings or in local organisations play an important role in this respect (Petermann 2015: 118 ff.;Beißwenger/ Hanhörster 2019).
Based on quantitative and qualitative surveys in two neighbourhoods in the German city of Dortmund, the aim of this paper is to determine to what extent and in what form poor households in socially deprived neighbourhoods gain access to assistance through social contacts. The research question is thus: To what extent does the social mix in a neighbourhood and, more specifically, in a person's immediate surroundings have an influence on resource accessibility? We are well aware of cultural boundaries and their role in hindering social interactions. However, ethnicity was not intended to be a main research focus here. Due to our interest in the role played by resource transfers in reducing the extent of social disadvantage -and potentially even in promoting upward social mobility -this paper explicitly focuses on individuals' social status and consequently on social boundaries and/or boundary-crossing.
In the following, we briefly review the research literature on different aspects of resource transfer among deprived population groups (Section 2). We continue by presenting the two neighbourhoods (Section 3) and the results of our surveys (Section 4). In the last section, we provide a short discussion and conclusion of our findings (Section 5). umgebung stellt sich -insbesondere für Familien mit Kindern -als wichtiger räumlicher Kontext für Kontakt und Ressourcentransfer heraus.
Schlüsselwörter: Sozial benachteiligte Quartiere; Ressourcentransfer; persönliche Netzwerke; nähere Wohnumgebung; soziale Mischung 2 Resource transfers as a key aspect of social integration Theories on social capital (i.a. Bourdieu 1983;Lin 2001) see interactions between residents and resource transfers as factors positively influencing social integration. Social interactions gain in relevance through swapping information, providing help in whatever form or just giving emotional support, and thereby transferring resources (Granovetter 1973;Lin 2001;van Eijk 2010). Such resources can help poor people to cope with everyday life and may possibly even support social mobility (Barr 1998;de Souza Briggs 1998;Forrest/ Kearns 1999;Forrest/Kearns 2001;Blasius/Friedrichs/ Klöckner 2008;Bailey/Besemer/Bramley et al. 2015). We should not forget that a person's available social capital is closely related to the size of his/her individual network and to the resources generally available through this network.
With regard to the kind of contact and the resulting resource transfer, Putnam (2000) and Narayan (1999) distinguish between 'bonding' and 'bridging'. 'Bonding ties' refer to ties within a social group, while 'bridging ties' refer to ties between different social groups. In scholarly research, these different types of contacts are closely linked with different forms of resources. Barr (1998) for example refers to 'bonding ties' as ties helping people to 'get by', i.e. help in overcoming day-to-day problems. By contrast, 'bridging ties' are often associated with social mobility, generating 'getting ahead' resources (de Souza Briggs 1998). Transcending social boundaries, such bridging ties are supposed to have a special role with regard to the social integration of disadvantaged sections of the population (cf. Farwick 2009: 258 ff.).
Looking in greater detail at these different ties, this contribution provides empirical evidence on the role bonding ties can play not only for getting by, but also in some cases for getting ahead. Furthermore, while scholars commonly refer to 'getting-by' and 'gettingahead' resources as two distinct categories, our research shows that the line dividing the two is often blurred (see also Blokland 2017: 36).

Resource transfers among people living in poverty
Network studies argue that, compared with the rest of the population, deprived people have relatively small networks offering little support and assistance (Lin 2001;Petermann 2015: 102). This is grounded in the fact that maintaining social ties costs money, something generally out of the reach of poor people (Andreß 1999: 161;Friedrichs/Blasius 2000: 63;Lin 2001: 65). A further explanation is that poor people often stay in the shadows, concealing their precarious financial status (Böhnke 2008: 135).
With regard to the types of contacts, many studies clearly show that bonding ties remain limited to the disadvantaged milieu and therefore are not very helpful when it comes to looking for work or an apprenticeship (cf. Friedrichs/Blasius 2000: 65 f.). Such marked bonding can be seen as a result of status homophily (McPherson/Smith-Lovin/Cook 2001), the principle whereby social ties mainly develop among people of the same (e.g. social or ethnic) group, as it can be assumed that they will have very much the same level of available resources (allowing reciprocal resource transfers) and similar values (cf. Farwick 2009: 169 ff.). In this context, a number of studies highlight the negative impact of a neighbourhood's ethnic diversity on social capital, seen to reduce either the quality of contacts between neighbours or trust in the neighbourhood or even both (Leigh 2006;Lancee/Dronkers 2011;Gijsberts/van der Meer/Dagevos 2012). We are thus well aware of the additional role played by ethnicity -alongside classin promoting or hindering social ties. Moreover, we are acquainted with previous research on how class and race interact in different contexts (Byrne 2009;Bilge 2010;Gillborn/Rollock/Vincent et al. 2012). However, in light of Letki's (2008) study on social cohesion in British neighbourhoods, demonstrating that a neighbourhood's socio-economic status impacts social capital significantly more strongly than ethnic diversity, this paper focuses solely on the social dimension of interaction and resource transfer. In this paper, therefore, bonding and bridging ties are exclusively associated with social difference, i.e. between people dependent on social welfare versus those in work.
Due to their high proportion of bonding ties, poor people are particularly reliant on forms of assistance generally available within their social group. These are for the most part 'getting-by' resources, for example emotional support, minor assistance or favours (Petermann 2015: 104). These groups thus miss out on bridging ties, i.e. ties giving access to people with new information and resources ('getting-ahead' resources) (van der Gaag/Snijders 2004;Pinkster 2007;Pinkster 2009;Pinkster/Völker 2009). Other studies show that poor households have much more local ways of life due to their limited resources, in many cases not having a job to take them out of their home surroundings (cf. Fischer 1982;Herlyn/Lakemann/Lettko 1991: 132 ff.;de Souza Briggs 1997;Friedrichs/Blasius 2000: 63;Small 2007;Blasius/Friedrichs/Klöckner 2008: 94;Pinkster/Völker 2009: 232). As a result, local networks of relatives, friends and acquaintances tend to dominate.
In light of the neighbourhood's significant role in structuring day-to-day life in poor households -in particular in comparison to other social groups -, the following sections focus on the spatial dimension of social interaction and resource transfer.

Effects of disadvantaged neighbourhoods on resource transfers
Early qualitative studies considering poor people's networks in disadvantaged neighbourhoods point to limited resource transfers among residents, with social networks playing no great role. Despite a high level of interactions, contacts are mainly superficial. In combination with the low level of resources of these sections of the population, the networks offer little reciprocal help (Herlyn/Lakemann/Lettko 1991: 123  Quantitative studies looking specifically at the influence of a neighbourhood's social mix on the existing types of contacts (bridging or bonding ties) or potentially available resources ('getting-by' or 'getting-ahead' resources) are however scarce. In one of the few studies available, Friedrichs and Blasius (2000) analysed the influence of a neighbourhood's social mix on the size of a person's support network and its local embeddedness. The study revealed that the higher the proportion of poor households in the neighbourhood is, the smaller is the size of a person's support network (Friedrichs/Blasius 2000: 66). Looking at the local embeddedness of the support network, the proportion of people experiencing poverty in the neighbourhood only had a significant effect in combination with education-related characteristics: the higher the proportion of poor households in a neighbourhood and the lower the level of education, the higher the proportion of local contacts was.
In a further study conducted in the Netherlands, Pinkster and Völker (2009) looked at the influence of a neighbourhood's social mix on the extent and type of available resources, clearly showing that access to resources helpful for finding a job ('getting-ahead' resources) was much more locally embedded in a poor neighbourhood than in one of mixed social composition (Pinkster/Völker 2009: 237). Overall, access to people with all kinds of jobs via local network contacts in a poor neighbourhood (controlling for different individual characteristics) was significantly lower than in a mixed neighbourhood. However, with regard to minor assistance ('getting-by' resources), a neighbourhood's social mix seemed to have little effect (Pinkster/Völker 2009: 238). These findings thus confirm the positive effect of a social mix in a neighbourhood on the transfer of 'getting-ahead' resources.
However, a survey conducted by van Eijk (2010) came up with different results. Again controlling for various individual characteristics, she was unable to find any influence of social mix in any of her three surveyed neighbourhoods on the proportion of residents with at least one local network person or an above-average proportion of local contacts in the support network (van Eijk 2010: 98 ff.). Similarly, with regard to the level of local contacts to well-educated residents (bridging ties), she could not find any distinct effect of the various levels of social mix in the three surveyed neighbourhoods within her multivariate regression analysis models (van Eijk 2010: 105 ff.).
Findings concerning the influence of a neighbourhood's social mix on the extent of local embedding of the support network and on the transfer of 'getting-ahead' resources are thus somewhat contradictory. It should however be pointed out that all of the studies described above look solely at a neighbourhood's social mix, without taking into account the heterogeneous nature of social mix at lower spatial scales.

The influence of structural-spatial crystallisation points on resource transfers
Alongside a neighbourhood's social mix, structuralspatial crystallisation points can also have an influence on encounters, the level of interaction, and the associated resource transfers (cf. van Eijk 2010: 43 f.). Feld (1981) uses the term 'foci' in reference to these crystallisation points, these hotspots of social interaction, describing them as "a social, psychological, legal, or physical entity around which joint activities are organized" (Feld 1981(Feld : 1016. Urban neighbourhoods generally provide a wide range of foci offering settings for contacts and in turn opportunities to build up and leverage social capital (Wellman 1996;Amin 2002;Völker/Flap/Lindenberg 2007;Wessendorf 2014). Foci can have either a formal or informal character, attracting different target and contact groups (Feld 1981(Feld : 1016van Eijk 2010: 43 f.). Neighbourhood foci include schools and playgrounds, but also voluntary organisations or even a community of residents in a single block of flats (Small 2009). Amin (2002: 969) uses the term "micro-publics" when describing the function of certain spatial structures in dealing with different forms of diversity: "settings where engagement with strangers in a common activity disrupts easy labelling of the stranger as enemy and initiates new attachments". The immediate housing environment can serve as a potential resource for establishing such micropublics by allowing regular encounters and meaningful exchange. Especially in exceedingly diverse contexts, these opportunity structures acquire a special function, facilitating the coexistence of different social or ethnic groups and allowing more social capital to be built up (Petermann 2015: 121).
In sum, the findings of the studies presented above reveal little systematic evidence regarding resource transfers among poor people in disadvantaged neighbourhoods or regarding the effects of social mix. Though many studies consider the effects of social mix on various indicators associated with the status of neighbourhood residents, none of them look explicitly at its effects on social interactions and resource transfers. The few studies looking specifically at the extent of resource transfers in disadvantaged districts, however, take a whole neighbourhood as the reference area, without zooming in on more local settings and their small-scale social mix.

Analysis of resource transfers in two deprived neighbourhoods in Dortmund
The following analyses look at the situation in the German city of Dortmund, a city whose internal structure is greatly affected by social-spatial polarisation (cf. Stadt Dortmund 2016). The two survey districts selected are characterised by both poverty and a high proportion of residents with a migration background.
The Hafen neighbourhood is a traditional inner-city working-class neighbourhood dating back to the 19th century, whereas Scharnhorst-Ost is a large estate on the outskirts of Dortmund built in the 1960s and characterised by a high level of (former) social housing. With its 18,000 inhabitants, the Hafen district is significantly larger than Scharnhorst-Ost with its population of 12,000. Poverty rates -measured as the share of residents living on welfare benefits -are similar in both neighbourhoods (Hafen: 35%; Scharnhorst-Ost: 32%) and well above the city average of 18%. Both neighbourhoods are also characterised by a significantly higher proportion of residents with a migration background than the city average of 34%: 66% in Hafen, and 64% in Scharnhorst-Ost (Stadt Dortmund 2016).

Methodology
A mixed-method approach was used for the study, made up of a quantitative questionnaire-based survey of residents and qualitative interviews. Both targeted poor residents (defined as people drawing welfare benefits) aged between 16 and 64. Focusing on welfare recipients not only implies that respondents have extremely low disposable incomes. Given that welfare recipients are not integrated into the labour market, they are additionally at high risk of being socially excluded and isolated from networks of employed individuals able to provide job-related information or other useful 'gettinghead' resources (van Eijk 2010: 101).
The quantitative survey was conducted face-toface with residents in both neighbourhoods, either at their front-doors or in their dwellings. Due to the small-scale heterogeneity of the social structure within the neighbourhoods, increased attention was paid to selecting respondents living in settings with varying degrees of social mix. Within these surroundings, the final selection of respondents was undertaken at random. In the case of multi-person households, the person contributing most to household finances in the past was the one questioned.
To gain more detailed information on the respondents' support networks and the resources they made available, a so-called resource generator was used (van der Gaag/ Snijders 2005;Petermann 2015: 163). This allowed us to determine the extent to which respondents were able to get various forms of assistance from their social networks in their daily lives. The respondents were asked about individuals able to give different kinds of support: minor favours (e.g. borrowing food), major help (e.g. helping to repair something), emotional support (getting by), or 'getting-ahead' resources, such as tips/help for finding a job/a new flat, filling out official forms or checking job application documents. The socio-demographic characteristics of these individuals were then captured and differentiated into bonding (i.e. contacts to other residents drawing welfare benefits) and bridging ties (contacts to residents in work).
Due to the high proportion of Turkish residents in the Hafen district, the questionnaire was translated into Turkish and interviews conducted by a native Turkish speaker. A total of 193 questionnaire-based interviews were conducted, 100 in the Hafen district and 93 in Scharnhorst-Ost. 56% of interviewees were women. The average age was 38. Interviewees came from 36 different countries (mainly Turkey, Syria and Poland) and 61% had a migration background. 32% of interviewees were single, while 28% lived with a partner and children. With regard to educational background, the lowestlevel school-leaving certificate was the most common in our quantitative sample (31% of respondents). Median residency in the neighbourhood was 5 years for the whole sample.
The qualitative study consisted of a total of 74 semi-structured interviews lasting between 30 and 90 minutes. Interviewees were mainly recruited in the neighbourhood social institutions, though a few were contacted during the questionnaire survey. Reflecting the ethnic diversity of the surveyed neighbourhoods, the interviews were conducted in seven different languages (German, English, Turkish, Arabic, Polish, Spanish and Portuguese). The respondents came from 21 different countries, with Turkey, Poland, Morocco and Syria as the main ones. Interviewees were aged between 20 and 65. The majority (74%) lived with a partner, and many of them had children. Thus, the qualitative sample is characterised by family households to a much greater extent than the quantitative one -an aspect needing to be kept in mind in the subsequent analysis.

Empirical findings
The following section presents -thematically structured -the empirical findings of the quantitative survey, supplemented by the main results of the qualitative interviews. It provides a deeper insight -over and above a purely quantitative view -into the everyday context of the resource transfers of poor residents in the surveyed neighbourhoods.
Extent and type of the support network One focus of our analysis was the extent to which the respondents received different kinds of support and assistance (minor favours, major help, emotional support, information). The findings of the quantitative sample on the size of support networks are to be found in Table 1. Only 4% of respondents were unable to name anyone able to provide support, pointing to a specific instance of social isolation resulting, according to the respondents, from a variety of causes, including not knowing anyone, preferring to deal with personal affairs alone, mistrusting others, or not having sufficient command of the language to interact with others. The other respondents had support networks with 5 contacts on average.
Taking account of the respondents' various characteristics and the different structures of the two surveyed neighbourhoods, the figures in Table 1 show that people with low levels of educational attainment have on average a slightly smaller support network. Moreover, the data reveals that in particular older respondents (over-50s), couples with children, other multi-person households, and to some extent male respondents and respondents with a migration background run a higher risk of not receiving any support or assistance from their personal networks. These findings are for the most part in line with those of previous studies on the extent of the support networks of poor residents. What is noteworthy, however, is the fact that the distinct characteristics of the two surveyed neighbourhoods with regard to the structure of foci have no significant influence on the size of support networks, even though the Hafen district (see Table 1) features a somewhat lower share of people without any support network at all.
Looking at the type of contacts in the support network, our analyses indicate that family ties were very important, with 30% of respondents stating that their support networks were made up mainly of family members. In 7% of cases, support networks consisted solely of family members.
Differentiated by bonding ties to other welfare recipients and bridging ties to individuals in work, our interviews provide no confirmation of the assumption drawn from previous research findings that most contacts belong to the bonding category. While the shares of respondents whose networks are made up for the most part or solely of people drawing welfare benefits are 20% and 7% respectively, a significant proportion (41%) of respondents, however, maintain no contacts at all with other welfare recipients. Our analyses thus indicate a fairly high proportion of bridging ties. Just 12% of respondents have a support network without any employed person in it. Nearly half (49%) have a support network made up for the most part of people in employment, and 14% even have a network made up solely of such contacts. Contradicting previous research findings, a large proportion of welfare recipients were thus found to have a network rich in ties transcending social boundaries.
With respect to the extent of local embedded support networks ( Concerning individual characteristics, we see clearly that singles in particular and, to a lesser extent, single-parent families seem not to receive any support or assistance from contacts in their immediate surroundings. By contrast, the probability of having one's whole support network made up of people living in the immediate surroundings was much higher among people in the 40-49 age bracket and in particular among over-50s. Age thus seems to play a key role in a support Migration background (n=193) Yes 5  6 No 1 5

Study area (n=193)
Hafen 3  5 Scharnhorst-Ost 4 6 network's local embeddedness. The networks of couples (with or without children) and to a lesser extent of people with a basic or intermediate school-leaving certificate are more likely to contain an above-average proportion of contacts living in the immediate surroundings. The figures indicate no significant differences with regard to migration background. Within the qualitative sample, welfare recipients' support networks also varied greatly in size. Nevertheless, nearly all were able to call up support and assistance for everyday problems. Generally speaking, the narratives point to the relevance of local support networks. In contrast to the findings of the quantitative analyses, the majority of contacts significant for resource transfers lived in the neighbourhood where the interviewees spent most of their time. Moreover, the qualitative sample additionally pointed to differences in the importance of the neighbourhoods, with Scharnhorst-Ost for instance having a much greater relevance in the residential biographies of the interviewees than the Hafen district: many interviewees had been born in Scharnhorst-Ost or had returned there after spending a few years living elsewhere.
The immediate surroundings, i.e. the stairwell or the block of flats, constitute a key crystallisation point for encounters and the subsequent transfer of resources. This finding is not out of line with the quantitative analyses, even if the latter relativise the importance of small-scale local networks. Looked at more closely however, we find that the dominant group of couples with children in the qualitative sample are more likely to maintain contacts with people in their immediate surroundings.
The interviews clearly show that the structuralspatial features of the immediate surroundings have a decisive role to play in establishing local support networks. Comparing the two neighbourhoods, we find that spatial structure is conducive to enabling encounters and contacts, thereby decisively contributing to resource transfers between residents. Contrary to the findings of previous studies, in which lower contact densities are attributed to large housing estates, the clear structure of Scharnhorst-Ost with its low-rise blocks of flats seems to increase residents' mutual awareness. Surrounded by these blocks of flats, the playgrounds and benches of Scharnhorst-Ost are used by all residents, allowing numerous opportunities for encounters and resource transfers, especially among households with children.
"I've got a bench right next to the front door of my block.
[…] Very practical when you get back from shopping: put your bags down, sit down and have a chat.
[…] the refugee family up there on the top floor, the wife was in the last weeks of her pregnancy, she often sat down there. And that terrible woman from next door with her walking frame, she would also sit down there. We used to sit there together in summer and sing songs with the kids. […] They're not people I know, but we see each other here and there. And then we find ourselves sitting outside on the same bench and pass the time of the day". For Susan, easily accessible benches are very important, as her restricted finances are not conducive to visiting cafés. The high number of opportunities to meet people in Scharnhorst-Ost with its much closer-knit community feeling is a great contrast to the Hafen findings.
One further surprising finding of the qualitative interviews is that a large slice of the support and assistance provided outside personal networks in both neighbourhoods comes about through chance encounters in the immediate surroundings. Such encounters involve people with whom the respondents have no closer (emotional) ties, but whose support and assistance can be readily called upon.

Types of support ('getting-by' or 'getting-ahead') by type of contact ('bonding' or 'bridging')
We now come to the core aspect of our study: the extent to which different forms of support and assistance are gained from these networks, with a distinction made between 'getting-by' and 'getting-ahead'. Previous studies show that the networks of poor people are generally used to transfer 'getting-by' resources, with little empirical evidence of transfers of 'getting-ahead' resources via bonding ties. With reference to the 'getting-by' resources, Table 3 first shows that 14% of respondents of the standardized interviews have no access to minor assistance, 21% to no major assistance, and 15% to no emotional support. This lack of access to 'getting-by' resources thus poses a major social problem. By contrast, a significant majority of respondents seem to have a secure supply of 'getting-by' resources.
In particular respondents without a school-leaving certificate are much more likely to have no access at all to minor assistance. As regards major assistance, respondents with contacts in the neighbourhood as well as respondents under 30 show a much lower risk of not receiving any support or assistance at all. Looking at emotional support, female respondents run a much lower risk of not being able to get such support. This contrasts with the higher risk prevailing among the over-50 cohort. With regard to migration backgrounds, the figures point to generally lower access to 'getting-by' resources for migrants and their descendants, especially for receiving minor assistance. a) = minor assistance; b) = major assistance; c) = emotional support; d) = formal assistance (e.g. help in filling out forms); e) = help in finding a new place to live; f) = help in finding a job Turning to access to 'getting-ahead' resources, Table  3 clearly indicates that the proportion of respondents with no access at all to such resources is significantly higher than for 'getting-by' resources: 43% of respondents have no access to help in filling out forms and dealing with the authorities, 47% have no access to help in finding a new place to live, and 66% have no access to help in finding a job. Thus, at least with regard to access to 'getting-ahead' resources for filling out forms or finding a new place to live, it would seem that more than half of the respondents actually do have access to such. The one area where this is definitely not the case, however, is in finding a job. These results therefore generally confirm previous research findings, illustrating that poor residents have greater access to 'getting-by' resources than to 'getting-ahead' ones.
The figures show further that especially respondents with contacts in their immediate surroundings run a significantly lower risk of not having any access to help in filling out forms and dealing with the authorities. Such local ties thus seem to play a major role in this respect. As regards help in finding a new place to live, it seems that the under 30s run a much lower risk of not receiving any help. Concerning help in finding a job, those respondents who only receive welfare benefits for a short period run a lower risk of not having any access to help. Moreover, we find a much lower risk -in the sense of an interaction effect -among people with support contacts in their immediate surroundings, especially when these are characterised by a social mix, i.e. immediate surroundings with a relatively low level (under 30%) of people drawing welfare benefits (in contrast to segregated surroundings with a welfare recipient rate exceeding 30%). 1 This finding points to the positive effect of social mix in the immediate surroundings on resource transfers, and especially transfers of those 'getting-ahead' resources so important for social mobility. A more detailed analysis of the effects of social mix is presented below.
Taking a closer look at the extent to which 'getting-by' and 'getting-ahead' resources can be provided in one way or another by either bonding or bridging ties, Table  4 shows that, with regard to 'getting-by' resources, there are no substantial differences between bonding 1 To calculate the share of welfare recipients in the respondents' immediate surroundings, the figures for the neighbouring blocks of flats were also taken into account (alongside those for the block in which the respondents lived), albeit with a lower weighting. This takes account of the view adopted by us that the immediate residential surroundings of the respondents refer not only to their own block of flats but also, to a somewhat lesser extent, to adjacent ones. and bridging ties. Only minor assistance is supplied to a somewhat higher extent by bonding ties. With regard to 'getting-ahead' resources, the figures show -in accordance with the findings of previous studies -higher shares of being supplied via bridging ties. The differences refer in particular to forms of formal assistance and help in finding a job. However, these dissimilarities are not as pronounced as could be assumed from the previous research literature. In this respect, it becomes apparent that bonding ties are also able to provide 'getting-ahead' resources to a certain extent.
The qualitative interviews additionally illustrate that forms of minor assistance, like lending food or tools ('getting-by' resources), are readily provided in most blocks of flats, even if contacts are little more than a quick 'hello' on the stairs. More complex forms of assistance, like driving someone somewhere or helping someone repair something, are also readily available in the immediate surroundings, as illustrated by the following quote: "He [a neighbour] was fiddling around with his car. I stopped to chat with him and then started talking about my car, where the speedometer wasn't working. And he told me he knew someone who could help me. So off we went to this garage. As a woman, you've got to watch out. There's always a good chance you'll get taken for a ride when you go to a garage".
The results of the qualitative analyses thus confirm good access to 'getting-by' resources, though such chance encounters do not always lead to a closer network tie. Similarly, one-off assistance is not always a precursor to regular assistance.
While 'more impersonal' forms of everyday minor and major assistance and support may be given by neighbours and chance contacts, family members are often the ones (initially) providing emotional support. As already made clear in the quantitative analysis, family members -insofar as they live close by -play a key role in the support networks of our interviewees. Their importance even increases when people, due to language problems, have difficulties building up social networks. The familiar family context makes it a lot easier to request support, even for those with a clear desire for self-reliance and independence or whose statements reveal a certain form of pride.
Alongside the many and often quite complex forms of 'getting-by' support and assistance, most interviewees -as already seen in the quantitative analysis -have only limited access to 'getting-ahead' resources. Only a few of them could report on people within their personal networks able to help in finding a job, writing job applications or providing information on the labour market.
The restricted access to 'getting-ahead' resources seems to be primarily associated with the lack of bridging ties, with only a few interviewees naming people in their networks able to provide such resources. In many cases, their contacts are themselves unemployed. In contrast to the quantitative analysis, however, bonding ties tend to dominate. Even when bridging ties do exist, they often seem unable to provide any additional information on how to deal with the authorities. Even contacts with higher social status do not always seem to have greater cultural and/or social capital from which the interviewees can benefit.
What is interesting is that both the qualitative interviews and the quantitative survey show thatcontrary to the initial assumption and to the usual use of the term -'getting-ahead' resources are not exclusively transferred via boundary-crossing contacts. Quite a substantial slice of such support is given within a group, pointing to the currently underestimated importance of bonding ties for transferring 'getting-ahead' resources. In certain situations, it would seem that contacts within a group 'fit' better, being able to transfer more relevant information and resources in the sense of getting ahead.
"The other neighbours also come from Romania, they work together with my husband, at the same company. As lorry drivers.
[…] They got to know each other at work [a previous job as a construction worker]." Interviewer: "How did he find the job? Was he tipped off by a colleague?" "Yes, a Romanian. He used to work there [on a construction site]. And then I got talking with him [the neighbour and former colleague]. He's a good bloke. Then I asked the boss whether he needed anybody." People working in the same field, as is the case with Anca's husband and his former colleague, are much more likely to help others find a job in this field than would be the case with a contact working in a more academic field of work. Contacts to people of higher social status ('bridging ties') are thus not always sufficient to promote upward social mobility. Much more important are the right interfaces, ones offering effective support or assistance.
To a certain extent, however, bridging ties to people with higher social and in particular cultural capital are sometimes indispensable, as seen by the example of 23-year-old Junis (from the Hafen district) who explicitly asks for support from his bridging ties when it comes to job applications: "As I just said, I've got a mate who's doing his Abitur. And that's who I generally ask, because he's already written lots of job applications.
[…] I've written quite a few myself, but they looked as if they'd been written by a kid from kindergarten. He'd often say to me: 'That's a load of shit what you've written. That's nothing like a proper job application'. […] In such cases, we help each other out." In this specific situation, his bonding ties, characterised by a similar level of educational attainment, seem unable to provide the necessary support.
The effect of the social mix in the immediate surroundings on the local embeddedness of support networks a) = minor assistance; b) = major assistance; c) = emotional support; d) = formal assistance (e.g. help in filling out forms); e) = help in finding a new place to live; f) = help in finding a job In a final quantitative step, we look at the extent to which the social mix in the respondents' immediate surroundings influences the availability of 'getting-ahead' resources. The following analysis only covers ties within a respondent's immediate surroundings, i.e. ties for the most part based on spatial proximity (locality-based ties, cf. van Eijk 2010: 107 ff.). Local contacts within the family were not included. With regard to bridging ties within the local support network, Figure 1 indicates first of all that the extent of bridging -regardless of the level of social mix in the immediate surroundings -is generally fairly high. Only 40% of the residents in areas where the immediate surroundings are socially mixed and nearly half of the respondents living in segregated surroundings reported no bridging ties within their local networks. Moreover, it becomes clear that the level of local contacts to people in employment is significantly higher among welfare recipients living where the immediate surroundings are socially mixed. 40% of respondents with immediate surroundings that are socially mixed can count on three or more bridging ties in their support network, whereas the percentage of respondents naming three or more bridging ties living in segregated surroundings amounts to only 8%. Using a linear regression test to check the positive effect of social mix on the number of bridging ties (controlling for other characteristics), we find this confirmed (Table 5).
Our final analysis looks at whether the social mix in the immediate surroundings influences the extent of 'getting-ahead' resources. As shown by Figure 2, only few people have local ties able to provide information on finding a job. However, major differences appear when zooming in on specific contexts. While in segregated immediate surroundings 88% of respondents were unable to name any local contact capable of helping them find a job, this percentage dropped to 60% in socially mixed surroundings. Generally speaking, the number of network contacts able to give help in finding a job is higher in socially mixed surroundings. This positive effect of social mix is -even if not very significant -again confirmed by a linear regression test (see Table 5).

Discussion and conclusion
In the light of increasing social-spatial polarisation in European cities, the urgent question arises as to whether deprived neighbourhoods have an additional disadvantageous effect on the social situation of their residents. There is currently a lack of research into whether access to resources facilitating upward social mobility is limited in such neighbourhoods, and into the role played by the neighbourhood or, zooming in, by a person's immediate surroundings on resource transfers. Empirical findings in response to this question are as yet rare with regard to German cities.
The results indicate a relatively high level of contacts providing mutual support and assistance, with only a few respondents unable to name any 'helping hands'. Contrary to the findings of Friedrichs and Blasius (2000: 66), we were unable to find any serious limitations to the number of such helping hands in deprived neighbourhoods. Moreover, the quantitative analysis -in contrast to the findings of Petermann (2015: 231) and Gestring, Janßen and Polat (2006: 53) -showed no significant effects of differing institutional and spatial structures in the two neighbourhoods, Dortmund Hafen and Scharnhorst-Ost. Based on qualitative interviews, we found a compensatory effect of two contrasting processes in Scharnhorst-Ost, a social housing estate built in the 1960s. In line with the research literature, it was to be assumed that the shortage of infrastructure and institutions in this neighbourhood would lead to a lower level of contacts and thus to smaller networks (cf. Small 2009). This can however be compensated by spatial structures promoting contacts in a person's immediate surroundings.
Differentiating between the types of contact in a person's support network, our quantitative analysis showed -in contrast to the assumptions found in the research literature (cf. Friedrichs/Blasius 2000: 65 f.; van Eijk 2010: 104;Petermann 2015: 104) -a relatively low level of bonding ties, but a high level of bridging ties. The qualitative interviews, however, indicated a higher level of bonding ties and a lower level of bridging ties, especially where families were concerned. The high proportion of bridging contacts in the networks of the quantitative sample leads to the conclusion that especially in deprived neighbourhoods, people with fixed-term, often precarious employment contracts similarly belong to some extent to the same disadvantaged milieu as our respondents.
Both the quantitative and the qualitative surveys show that bridging ties are not necessarily associated with 'getting-ahead' resources. For instance, our quantitative survey revealed that, even when respondents had numerous bridging ties, they still did not receive many 'getting-ahead' resources. Our qualitative results pointed to the importance of the right 'interfaces' for transferring resources: for resources promoting upward social mobility to be transferred, it is not enough just to have

years and older
Ref. Ref.

Level of education
No grade -0.508 -0.062 Lowest-level school leaving certificate Ref. Ref.
Intermediate-level school leaving certificate Ref. Ref.
Highest-level school leaving certificate 0.518 0.010

Foreign diploma
Ref. Ref.

Duration of welfare receipt
Up to 12 months (short) 0.320 -0.073

months and longer (long)
Ref. Ref.

Level of social mix in the immediate surroundings
Socially well-mixed (below 30% welfare recipients) 1.394** 0.562* Segregated (30% and more welfare recipients)
Cases contact with partners well-endowed with such resources. A certain affinity between giver and taker is also needed. In addition to the 'getting-ahead' resources passed on by people, it seems also important to have low-key institutions enabling greater access to such resources (cf. Small/Jacobs/Massengill 2008). This is however a field where more research is needed.
The qualitative interviews also show that the often assumed -as in our research -demarcation between 'getting-by' and 'getting ahead' resources is a lot more complex than commonly suggested in the literature (cf. also the criticism raised by Bailey/Besemer/Bramley 2015: 297). The dividing line between the two often seems blurred, with any demarcation dependent on specific situations. For instance, the question is raised as to whether having someone to look after the children to enable participation in a language course is perhaps a form of upward social mobility, as language skills are a sine qua non for getting a decent job. A nuanced differentiation is therefore necessary here.
No strong local embeddedness in a person's immediate surroundings was perceived in the quantitative survey (cf. also Blasius/Friedrichs/Klöckner 2008: 94;van Eijk 2010: 95 f.). However, we also found that couples (with or without children) -accounting for a major proportion of respondents in the qualitative survey -tend to have many more people offering support and assistance who live in their immediate surroundings. This is where the mixed-method approach used by us bears fruit, allowing us to arrive at a differentiated view of local support in the form of personal networks and contacts. For instance, while the respondents in the qualitative survey attach greater importance to loose contacts in their immediate surroundings, the persons named in the quantitative survey via the resource generator tend to belong to the inner circle of the respondents' existing personal networks, many of whom do not live in the immediate surroundings. The high importance of such loose contacts and their ability to also provide resources going beyond just 'getting-by' resources points to the need for further research into the effects of these looser forms of contact.
Last but not least, our results indicate that the social mix of the immediate surroundings plays an important role in defining locally embedded support networks, with bridging ties being much more prevalent when the surroundings are well-mixed. For instance, respondents living in such surroundings are much more likely to receive help in finding a job. But here again, further research is needed to bolster the significance of this finding.