connected communities - how social networks power and sustain the big society

8/8/2019 Connected Communities - How social networks power and sustain the Big Society

http://slidepdf.com/reader/full/connected-communities-how-social-networks-power-and-sustain-the-big-society 1/83



5

We believe that understanding and utilising social networks shouldinform and can bene t any policy area. There are alreadydiverse examples: the Department of Health’s Putting People Firstprogramme, which aims to transform adult social care, explicitlyrecognises the importance of social capital and networks in helpingpeople to gain control over their own lives. 20 At the other extreme,military operations now seek to build support for their actions byin uencing local social networks through key individual hubs, andgather intelligence through online social networking. 21 Christakisand Fowler have illustrated the policy potential of social networksin addressing public health issues such as smoking and obesity, andraising voter registration and turn out.

Even policy areas that do not immediately appear to bene t fromnetwork approaches, such as reducing pensioner poverty, can augmenttraditional approaches such as tax adjustments and fuel payments.Connecting pensioners to richer, more supportive networks canreduce isolation and disempowerment, with the knock-on effects of reducing social care costs. Fostering informal exchanges and supportthrough networks, and the knowledge and opportunity gained throughstronger social ties, can lead to reduced spend by pensioners,increasing disposable income. These lines of thinking are critical inan age of austerity.

CHAllenges

While we are excited by the possibilities of visualising and usingsocial networks at a community level, we acknowledge that we are atan early stage in our project, and are still in the process of re ning ourexploratory research methods and building the evidence base of ourargument. Social networks give us powerful information about socialrelationships, but they tell us little about how individual psychologies orglobal markets impact on communities, and cannot offer a comprehensiveaccount in themselves.

In what follows, our empirical ndings, from quantitative data,qualitative research, case studies and social network analysis (outlinedin the methods section below) are integrated into the text as evidenceto support our overall narrative and argument. While we have strivedto build a strong evidence base for our claims, researching socialnetworks is an exacting process, and most of our ndings should beregarded as indicative and exploratory, rather than representative and

nal. In this report, we have deliberately selected parts of the analysiswe have undertaken to date in order to illustrate the points we exploreand put forward. Forthcoming RSA brie ng papers will excavate therich, complex data we have collected to analyse themes in more detail.

We are also realistic about what our approach might ultimately achieve.While we believe social networks offer a powerful tool that may wellenable communities to solve problems and shape circumstancesmore effectively, no social network can provide a substitute for capitalinvestment, or form the rationale for signi cantly withdrawingsupport and funding from areas where entrenched disadvantage is

acute. We view our work as an important and timely perspective ina climate of economic austerity, not as a panacea.

introduCtion

...social network interventions canattempt to create benign socialviruses, through which smallinterventions seek to create major

impacts through contagion effects



6How soCiAl networks power And su stAin tHe big soCiety

struCture

The structure of this report is as follows:

— s c outlines our perspective on the main themes of thereport, including communities and social networks. We contextualiseour approach with respect to the Coalition Government’s aimto build the Big Society in an austere economic climate. We alsoconsider the potential of social network theory and measurementto transform policy-making more generally.

— s c begins with some conceptual clari cations, and drawsout some important lessons learned from community policy inrecent years. The principal shift is a move from an area and de citbased model of communities to an emphasis on social assets, whichwe are developing by focusing on particular aspects of social networks.

— s c h explains our methodology, and provides thumbnailportraits of our action research sites: New Cross Gate, southeastLondon and Knowle West, Bristol.

— s c outlines the theory of social network structure andfunction, and examines each aspect of this theory in light of theemerging data from our eldwork in New Cross Gate and KnowleWest. It also discusses the potential bene t of social network analysisand re ection as a process which itself strengthens networks.

— s c velaborates our account of how this network perspectivecan help foster qualities communities need in the current economicclimate, like social inclusion, resilience and empowerment.

— s c is about the process of building connected communities,and using social networks to create social capital by placing moreemphasis on the role of social networking in helping to weave certainkinds of social networks that we believe to be desirable for communities.

— s c v attempts to address how our approach to communitycould inform the government’s vision of the Big Society.

— o A outlines plans for year two of the Connected

Communities programme.



7

seCtion 2: retHinking Community

We are all members of several communities and our ties with them canincrease or decrease. It is both illogical and dangerous to corral people asif they could only belong to one community.

— A a a s22

In this section we begin by making a distinction between two conceptsthat are often con ated: social networks and social capital. We do thisin order to highlight a distinct property of social networks – that theycan be visualised. Social network analysis can inform the optimal useof social goods (community assets which are publicly available butscarce and contested).

Subsequently, we draw some lessons from the modern history of community policy and practice, including the limitations of viewingcommunities in exclusively geographical terms, and we examinesome recent research on social assets that points to the need for adeeper understanding of the structure and function of social networks.

2.1 soCiAl networks, soCiAl CApitAlAnd soCiAl goods

‘Things should be as simple as possible, but no simpler.’ — A e

As a rough and ready heuristic, the concept of social capital is invaluable,

because it allows researchers to put into operation the obvious pointthat relationships (social) have considerable value (capital). Yet asan analytical device the concept is rightly contested, because there aremany kinds of relationships, and their value differs across contexts,in degree and in kind. Indeed, economist b f describes socialcapital as a ‘totally chaotic, ambiguous, and general category that can beused as a notional umbrella term for almost any purpose.’ 23

We accept that the concept of social capital is not precisely analyticallyde ned, but political scientist r p a, perhaps the mostfamous proponent of social capital, offers the following minimalistde nition: ‘Social connections and the attendant norms and trust.’24

The Connected Communities programme attempts to make sense of what is meant by ‘the attendant’ in this context. How do socialnetworks shape social norms at a local level, and can such norms, inturn, help to build trust such that informal transactions become easier,people feel safer, and more likely to help each other out, for instancewith child care, shopping, or gardening, in a way that is be tting of the coalition government’s vision of the Big Society?22 Sen A, i a V c : th i d, 2006,

Allen Lane.

23 Halpern D, s c a Ca a, 2005, p13, Polity Press. Forthe full critique See Fine B, th s c a Ca a :r a ch b hav ba, 2010, Pluto Press.

24 Putnam R, ‘Tuning in, Tuning out: the strangedisappearance of social capital in America’.Political Science and Politics, 1995 28: 1-20.

retHinking Community




tHe VAlue of soCiAl CA pitAl

th v a a h a ch. ba a c a ch a a

h j c ha , a h a c a a h c a h c a ac h

a ca a :25

Our friend lives at the end of the terrace and his next door neighbour, an elderly man, recently moved in, tripped and fell into his front garden. My mate came out and took him to thecafé for a cup of tea. Long story short, they now meet for teaeach week or so, he helps him with his garden, he’s plugged intothe lunchtime social specials they have there, the police safer neighbourhood lot meet there and now know him and look out

for him. They have weekend reading sessions for kids. And thecafé itself buys all its ingredients from the other local traders,and they’re completely into fresh healthy ingredients througha local network.

It’s all probably much more effective than anything the Council could deliver by itself. You could start to see it and encourage itas the hub of a network for social good.

But isn’t who goes there just opportunistic to an extent? If youcould identify who uses places like this café and why; how tomake them more inclusive and understand and address local problems effectively; how to identify and use these kinds of

resources in the rst place, well…It’s kind of social enterprisethrough reconceived existing networks. And if you could encourage other positive spin-offs, what good stuff could peoplewho meet there together do? That’s what you’re talking aboutisn’t it?



retHinking Community 9

Measuring and mapping existing social networks is a means to an endin building the kinds of supportive and productive relationshipsencapsulated in the term ‘social capital’. Indeed, while social networksand social capital are inextricably linked, social networks are importantin their own right. As a term, ‘social capital’ is widely used bysociologists, economists and politicians, and has to an extent becometheoretically saturated, while social networks can be measuredempirically and represented visually. We are just beginning tounderstand how powerful these patterns of connectivity might be asa tool for evidence-based policy and practice.

Social networks are therefore a necessary condition for social capital,and we focus on examining these necessary conditions, and sometimestheir absence, with a view to informing practical interventions. We careabout norms and trust, but if these phenomena are, as Putnamsuggests, ‘attendant’ on social connections, then there is an obviouscase for focusing on the connections themselves.

Such an approach is timely because social networks have becomeeasier to map and measure in recent years, due to the proliferation of telephone and internet data and increasingly sophisticated socialnetwork analysis software. These re nements in social networkresearch may allow us to understand communities as the emergentproperties of networks, in a manner similar to the way that thedevelopment of brain imaging allowed us to understand cognition asthe emergent properties of neural networks.

However, the technical language of social networks is nascent andcurrently abstract and unfamiliar, so we use a more familiar lexiconof social capital. We have been guided by political scientistdav Ha ’analysis of the concept, which recognises the rangeof meanings and applications given to social capital, but contends thatthey are part of the same ‘sociological genus’. This genus expressesitself through components (networks, norms and sanctions); levelof analysis (micro, meso and macro); and function (bonding, bridgingand linking). 26

In Halpern’s terms, the salient aspects of social capital in this reportare networks and norms at the meso (community) and micro(individual) levels, with an emphasis on bridging between groups andlinking to sources of power and in uence. We recognise thesecomponents and levels of analysis are not independent of otherbranches of the social capital framework, namely individualpsychologies, government policy or global markets, but we havechosen to narrow our focus in order to relate our network researchto future practical action.

One major challenge to the role of social capital in communityregeneration is that it is as likely to be used as a private asset forindividuals as a public good for communities. 27 At a recent seminar todiscuss the emerging notion of the Big Society, ba Q , CEO of the London Borough of Lewisham, distinguished between socialgoods, community assets which are publicly available but scarce andcontested, and public goods, which are genuinely a resource for thewhole community.

25 http://rsaconnectedcommunities.wordpress.com/2009/02/04/meso-soup /

26 Halpern D, 2005, op. cit

27 Ibid

http://rsaconnectedcommunities.wordpress.com/2009/02/04/meso-soup







Quirk argues that a major challenge for local authorities is tryingto ensure that places like community centres are public as well associal goods. In our research, Somerville Adventure Playground inNew Cross Gate emerged as an example of a social good that is notcurrently functioning as a public good due to network effects(see section 6.2) .28

When communities are viewed as large cohesive groups, a communitycentre may operate as a public good in the sense that one person’s useof it need not interfere with another’s use. However, whencommunities are viewed as places that feature con ict between groupswith competing aims or values, such public places become contestedsocial goods. Social network analysis is potentially valuable in helpingto make social goods less contested. It may, for instance, show thattwo groups that felt they were radically different may have commonmembers, or be connected to the other group through a common tie.In this way, new common ground and shared agendas are identi edas platforms for con ict resolution, mutual understanding, trust andpotential collaboration.

Ideas of social networks, capital and social and public goods are key tore-imagining how we have come to formulate community policy andwork in deprived neighbourhoods, principally through supplementaryprogrammes such as area-based initiatives.

2.2 tHe rise And fAll of AreA-bAsed initiAtiVes

The design of area-based initiatives over the past few decades, fromCity Challenge (1992-1998) through to the Single RegenerationBudget (1994-2004), culminating in New Deal for Communities(NDC, 2000-2011) has increasingly placed communities at the heart of the regeneration process. The rationale for the geographic conceptionof community that underpinned these initiatives stemmed partly fromNew Labour’s emphasis on the devolution of power, together withrecommendations by Policy Action Teams that if communities ledthemselves, they would be better placed to drive participation, localownership and capacity building.

Longitudinal data 29 on the impact of NDCs shows absolute improvementson around 80% of performance indicators, across community safety,education, health, and physical environment. 30 For instance, there has

been a 7% increase in people’s perceived ability to in uence decisionsthat affect their local area in the 2002-2008 period and a 12% increasein people feeling part of the community, but, notably, no signi cantincrease in people taking part in education and training, and noincrease in employment. 31

More generally, the NDC model did not always work as policy makershoped it might, and the approach has suffered from diªculties thathave dogged the area-based approach for years, particularly unrealisticexpectations about what can be achieved, shallow and unrepresentativelocal engagement, and the periodic burnout of a small core of community activists.

Social network analysis is potentiallyvaluable in helping to make socialgoods less contested.



11

One perplexing nding to emerge from the recent data is that inNDC communities, at an aggregate level, indicators of social capitalhave failed to improve beyond the increases experienced by othersimilarly deprived neighbourhoods. 32 Given NDCs’ comprehensiveapproach, and the research evidence on the importance of socialcapital in affecting economic performance, crime, health andeducation – all NDC priorities – this is a challenging result. 33

One possible explanation is in the limitations of how social capital hasbeen measured. In this context, it was based on individual responsesto questions on trust in local institutions, feelings of belonging, andthe extent to which people value and trust all other local people takenas a whole. It excluded the value of social relationships throughnetworks, which we believe is more critical.

Additionally, a 2010 Department of Communities and Local Government(CLG) report suggests that possible explanations lie in ambiguity inthe aims of resident participation and the fact that for many residents,their experience of participation was not a wholly positive one. Also,not enough residents engaged in NDC activities to generate signi cantimpacts on social capital at the community level. 34

We believe that a further and more fundamental explanation lies inthe limitations of approaching communities in solely geographicalterms. Much of the existing approach is contained within speci carea-based initiatives (ABIs) – publicly-funded pilot programmestargeting areas of social or economic disadvantage. These programmesaim to improve residents’ quality of life across multiple outcomes,including employment, skills development, health and crime. But inan overcrowded regeneration landscape cluttered with programmesand agencies, there is a danger of initiative and participation fatigueand a perception that resident initiative and interaction requiressome form of ‘professional’ public or third-sector brokerage, ratherthan succeeding purely on an informal and self-organised basis.

Some area-based initiatives with tightly-de ned and protectedgeographical boundaries can reinforce isolation and limit the socialresources available to that area. Such boundaries create strong internalidentities but weaken the ability and willingness to connect to anddraw support from other areas. ABI rules that determine who canbene t from expenditure and who can participate in decision making– namely ABI residents – reinforce these divides. As a result,communities can fail to build the bridges across social and economicdivides that have been shown, particularly through Americansociologist ma g a v ’work on the value of ‘weak ties’, toimprove economic performance. 35 For instance, we know that if thenumber of those unemployed in an area exceeds around one in four,the likelihood of escaping poverty drops dramatically. 36

ABIs rarely have enough time and resources to understand the socialnetworks that shape how the communities function. While theNeighbourhood Renewal programme helped to drive the increasedavailability of small area statistics, this often took the form of tablesof quantitative data displaying de ciencies against local or nationalaverages: unemployment and crime rates are high; educationalattainment and health indicators are low, and so on.

28 Barry Quirk, presentation at Big Society Public ServicesSeminar: r v v ?Institute forGovernment, 6 July 2010.

29 Data collected on the same subjects at multiplepoints in time. For a full de nition see:http://www.caldercenter.org/whatis.cfm

30 Ipsos-MORI, na a eva a n d a C p a : H h s v da a,2002–2008 .

31 http://extra.shu.ac.uk/ndc/downloads/general/Volume%20two%20-%20Involving%20local%20people%20in%20regeneration.pdf

32 Communities and Local Government, ma va a ac v : ev c h n d a C p a th n d a Cna a eva a : f a r – V 3, 2010.

33 In some models, social capital is more important foreconomic growth than human capital (Whitely, ec cg h a s c a Ca a ,1997, p18, Political EconomyResearch Centre, University of Shef eld). Widespread trustplus shared informal norms and social habits reduce crime(Sampson et al, Neighbourhoods and Violent Crime: amultilevel study of collective ef cacy , 1997,sc c, 277,p918-24). The impact of social capital on health has beenknown since Durkheim’s work on egoistic suicide in 1897and the positive relationship between social capital andeducational attainment is indicated in Baron S, Field J,and Schuller T, s c a Ca a : C ca p c v, 2000,Oxford University Press. All sources traced via Halpern D,s c a Ca a, 2005, Polity Press.

34 Communities and Local Government, 2010, op. cit

35 Granovetter M, The Strength of Weak Ties, 1973, A caJ a s c , Vol. 78, Issue 6, p1360- 80, andGranovetter, M, The Strength of the Weak Ties Revisited,

1983,s c ca th , Vol. 1, p201-33.

36 Halpern D, 2005, op. cit.


http://www.caldercenter.org/whatis.cfm

http://extra.shu.ac.uk/ndc/downloads/general/Volume%20two%20-%20Involving%20local%20people%20in%20regeneration.pdf






http://www.caldercenter.org/whatis.cfm




Missing from the intelligence is the extent and strength of (ofteninformal) human relationships and exchange that form communitiesand that enable community regeneration in its broadest sense. Suchintelligence is fundamental. A lack of understanding can lead toa de cit model of neighbourhoods, ‘pathologising’ communities andassuming they need top-down ‘treatment’ through an ABI. While suchareas are objectively disadvantaged, understanding them in de citterms alone can disempower and stigmatise, and fail to recognise andmobilise the assets that do exist.

Immediate pressure to spend money in accordance with annual budgetcycles means there is limited opportunity to develop this understanding.There is also a tendency to build new infrastructure and deliveryarrangements on top of or alongside the old. These hurried interventionsoften yield no signi cant change in the ways that communitiesfunction, and people retreat to relatively closed networks, where groupmembers look to themselves rather than the wider community.

Such lessons are key if the Big Society and future attempts to addressproblems in deprived areas are to succeed.

2.3 CHAnging How Communities CHAnge

Our reading of the available evidence on community policy is thatfor the last three decades governments have served to perpetuatewhat pa wa z a ccalls ‘ rst-order change’ – changes made withina conventional frame of meaning that tend to be implemented whenthe overall approach to social change has not been fully thoughtthrough, is adjusted incrementally, or has not been adaquately updated.

Addressing the problems of eªcient resource allocation, unlockinghidden wealth, and fostering pro-social behaviour outlined aboverequires a more radical ‘second-order change’— a shift in the frame of meaning. Community policy needs a different concept of ‘community’in order to affect the manner in which things change. Our attempt tounderstand communities as social networks is an attempt to generatethis more fundamental model of change. 37 In attempting to shift theway society operates — in re-thinking the social contract and how weprocure and produce services — the emerging notion of the Big Societycan similarly be seen as an attempt to create a second-order change.

As community development consultant A g chargues, peoplepartake in various kinds of communities for a variety of purposes, andspan communities of:

• identity, to share cultural activities;

• interest, to pursue shared fates;

• purpose, to achieve common goals;

• practice, to exchange experience and learning;

• inquiry, to investigate an issue collectively;

• support, to provide mutual aid; and

• circumstance, to deal with unexpected situations.

While such areas are objectivelydisadvantaged, understanding them in de cit terms alone candisempower and stigmatise, and

fail to recognise and mobilise theassets that do exist.



13

What these different forms of community share is that that they areall better understood as networks rather than geographical or interestareas, but networks which nonetheless differ fundamentally in theirstructure and function in ways that should inform any attempts to‘develop’ them.38

Our appreciation for the relationship between social networks andcommunity improvements needs deepening. As sociologistA C a succinctly puts it, ‘geographical propinquity does not meansocial communality’ .39 Geography can dictate community, particularlyon housing estates where a whole range of issues are common to allwho live there, and local authority boundaries and shared residentialareas afford some common ties, but there are many other connectionsbetween people that may be underappreciated, at least partly becausethey remain invisible.

Those not willing to engage in the local safer neighbourhood forum,for example, may take part in social clubs, have children at the localschool, share communal gardens and so on. Despite lack of awareness of, interest in, or cynicism towards formally organisedmeans of addressing social problems, our research indicates that most‘disengaged’ people do have social ties within their communitiesthrough a variety of channels. Such ties are important because theyprovide leverage points to improve people’s lives, but at present theyare rarely acknowledged and seldom used.

Research produced by Communities and Local Government in thelast few years has increasingly recognised the importance of suchrelationships. For instance, a recent working paper suggested that‘Community ties, social reciprocity and civic engagement are particularlyimportant to addressing contemporary health, skills and environmentalchallenges.’ Six key relationships that need to be better understoodwere identi ed:

• Horizontal relationships; between friends and neighbours, betweenpeople and public spaces (public order and civility) and relationsacross people of different backgrounds (community cohesion).

• Vertical relationships; between service providers and service users(co-production), participation in voluntary activity, and participationin civic and political activity. 40

All of these relationships are best understood in network terms, butthe main value of social network analysis is that it allows horizontalrelationships to be measured in a way that may in uence public policy.

At a roundtable discussion on deprived neighbourhoods in January2010, participants agreed that one of the main building blocks inregeneration is recognising that deprived neighbourhoods are not allthe same and therefore require tailor-made interventions. Mappingsocial networks is certainly part of making tailor-made, network-sensitive interventions work. For instance, there is good evidence of the impact of peer effects and social norms across a range of outcomes, and also evidence that people in deprived neighbourhoodshave lower levels of trust and bridging capital. 41

37 Watzlawick P, Cha : p c p f aa p r , 1974, WW Norton & Co.

38 Gilchrist A,th w -C c C (2 e ),2009, Polity Press.

39 Clark, Ar a l w pa : u aC , May 2007 http://www.reallifemethods.ac.uk/publications/workingpapers/2007-05-rlm-clark.pdf

40 Communities and Local Government, s c a Ca a aC v c pa c a : A s a c r v C

p c, 2010.

41 Communities and Local Government, 2010 op. cit


http://www.reallifemethods.ac.uk/publications/workingpapers/2007-05-rlm-clark.pdf







Such evidence is important because there is currently some debateover the extent to which the coalition government expects citizens to‘take over’ public services, and the success of such ventures will haveto be tailor-made, depending on the public service and thecommunities involved. A recent report from community consultants,Public Agency and Community Empowerment Strategies (PACES)argues that although communities cannot ‘take over’ public services,they can complement and co-produce them:

‘Most of the successful examples in this eld are actually sharing of both power and delivery by a public service and a community organisation.Tenant management organisations, friends of parks groups, police-residentliaison groups and many others perform this cooperative function.’ 42

How to nd the right people for these cooperative functions, so thatthey can become sustainable models of community engagement,rather than exceptions to the rule that depend on the efforts of selectindividuals, remains unclear. A social network approach to communityregeneration could also play a role in making more eªcient use of scarce nancial resources, because network analysis can, in principle,tell us which community hubs or members are best placed to have anoverview of existing community skills and needs, coordinate activityand spread useful information and opportunity. We might assume thatsuch people are elected representatives, but in New Cross Gate, forexample, we found that more people know and gain value from theirpostman than their local councillor.

Social capital and social networks have become explicit policy anddelivery aspirations in recent years. For example, the Department of Health’s vision, expressed in its ‘Putting People First’ policy paper, hasbeen to transform adult social care and recognises social capital asa key ingredient in enabling people to live their lives as they wish, withindependence, well-being and dignity. 43 Meanwhile, the national drugstrategy is recognising the importance of social capital in helpingusers seek and sustain recovery. The RSA’s work with drug users inWest Sussex and in Peterborough is piloting ways in which ‘recoverycapital’, in its personal, social and community forms, can be betterunderstood, grown and mobilised to help drug users maximise healthand wellbeing and participation in the rights, roles and responsibilitiesof society.

Existing examples of related policy research include the ‘Social AssetsResearch Report’ by the Community Foundation of Northern Ireland,in which 890 super output areas were given estimated scores for theirlevels of social assets, based on measures of bonding, bridging andlinking social capital and community capacity and capability. 44 The report nds that identifying areas in need in terms of multipleindices of deprivation should be complemented by a richerunderstanding of social assets, and that this understanding can leadto ‘greater sensitivity in funding allocations’, which is consistent withour claim that social network analysis can give a more meaningfulunderstanding of eªciency. It also reports that trust is a keyprerequisite for any successful community regeneration work.

A social network approach tocommunity regeneration could also play a role in making moreeªcient use of scarce nancial

resources, because network analysis can, in principle, tell uswhich community hubs or members are best placed to havean overview of existing communityskills and needs, coordinateactivity and spread usefulinformation and opportunity.



15

The NESTA-funded ‘Young Foundation Report on Social Innovation’uses a form of web-based organisational network analysis to comparenetworks and levels of networking connected to innovation, basedon identifying ‘hubs’, ‘gatekeepers’ and ‘pulsetakers’, and seeks todiscover the optimal levels of network density for fosteringinnovation. 45 Their report highlights that the local authority, alongwith schools and colleges appear to dominate the innovation processand that the potential for social network analysis to inform thestrengthening of communication networks remains largely untapped.

A recent IPPR report examines how Polish immigrants use socialnetworks to nd employment, and suggests that while this makesthem more likely to nd work in the short term, it can lead to socialstrati cation in the long term, because opportunities are restrictedto the range of the social network, which may cut them off frombarriers to integration, including language and knowledge of the widerlabour market. 46

These three sets of recent ndings relating to social assets with respectto resource allocation, innovation, and immigration respectively,provide some corroboration for our contention that communities needto be understood as networks, and also that a deeper account of socialconnectivity is required.

Summary

Decades of community policy and practice have indicated that viewingcommunities solely as geographical entities has limitations.In recent years there has been a growing emphasis on social assets,and an emerging understanding in the policy world that a deep graspof social assets is only possible by uncovering the nature of community connectivity.

42 Chanan G, and Miller C, th b s c : H Cw : A p v i a a r Ca ca, Spring2010, PACES,http://www.pacesempowerment.co.uk/pacesempowerment/Publications.html

43 p p f : A sha V a C h t a a A s c a Ca, 2007,

HM Government, London.

44 Morrissey M, Healy K, McDonnell B, s c a Ar a ch r : A n A ach u aa w h C , 2008, The CommunityFoundation for Northern Ireland and CommunityEvaluation Northern Ireland.

45 Bacon N, Faizullah N, Mulgan G, Woodcraft S,t a : H l ca A a i va acha c a , 2008, NESTA.

46 Sumption M, s c a n a p h i a h uk, 2009, Institute for Public Policy Research.


http://www.pacesempowerment.co.uk/pacesempowerment/Publications.html







seCtion 3: reseArCH metHods:unCoVering soCiAl networks

The RSA has attempted to measure social networks, and has chosento do so in an area that is coming to the end of an ABI (New Cross Gate,southeast London); and in an area that is geographically isolated, withfew public goods save for an innovative community centre, and thataccording to national statistics is multiply-deprived (Knowle West, Bristol).

We recognise that these areas are not universally representative, butsuggest that they house tensions, problems, and hidden assetscommon to many communities and in this way our work is illustrativeof approaches that may be replicated in other areas.

We describe below the methodological framework that has guided us,and outline the properties of networks that have shaped our choice of methods. We then offer a brief overview of our research sites, beforedescribing our research tools and processes in detail. The section endswith a discussion of the limitations of our approach.

3.1 metHodologiCAl rel Ationism

Taking social networks seriously means recognising that theelementary unit of social life is neither the individual nor the group.Social networks allow us to move beyond this classic theoreticaldistinction. This is because they presuppose a social structure thatboth shapes and is shaped by individual behaviour. As Christakis andFowler indicate: ‘The science of social networks provides a distinct way of seeing the world because it is about individuals and groups, and about howthe former actually become the latter.’ 47

This thinking underlies the RSA’s perspective on research moregenerally. Methodological individualism provides social explanationsbased on aggregating the signi cance of individual actions and formsthe basis of most economic theory, while methodological holismexplains individual actions on the basis of the social structures inwhich individuals are embedded and through which they are de ned. 48 Modern sociologists such as A h g recognise that structureand agency are interrelated and have a recursive character. 49 Ratherthan explain social phenomena through the relationships of structureand agency, it is possible to make social relationships themselves theprincipal units of analysis. This is the basis for methodologicalrelationism. In this framework, psychology and social networks worktogether and both need to be incorporated to understand individualbehaviour and social phenomena. Methodological relationism isrelatively rare in western social science, but is more common in Asiansocial psychology, in which the unit of analysis is characterised as follows: ‘…Not the individual or the situation alone, but person-in-relations(focussing on a person in different relational contexts) and person inrelation (focussing on persons interacting within a relational context)…

It acknowledges that the social ‘presence’ of others is always entered intosocial calculations…The process is bidirectional…it recognises theindividual’s embeddedness in the social network.’ 50



reseArCH metHods: unCoVering soCiAl networks 17

Presently, our Connected Communities project focuses more closelyon social networks than on individual psychology. As we progress thisyear and next we will seek to integrate our work with the RSA SocialBrain project, and begin to show the power of research grounded inmethodological relationism, which, as Robins and Kashima state:‘…directs researchers to theorise about an individual’s action whileincluding the workings of the individual’s egocentric local network in its

analytical scope.’ 51

3.2 ConneCtiV ity, ContAgion And reflexi Vity

The theoretical basis for social networks is well established, 52 andsocial network analysis has now become a burgeoning academic eldacross disciplines, in part facilitated by data trails made possible bynew technology. For instance, a recent large-scale study reported inScience combined the most complete record of England’s nationalcommunication network with national census data on the socio-economic well-being of communities, and concluded that the diversityof individuals’ relationships is strongly correlated with the economicdevelopment of communities. Such a powerful nding was onlypossible by using data from more than 90% of the mobile phonesin England. 53

The network perspective offers a distinctive explanatory tool becauseit reveals patterns of relationship and exclusion that would otherwiseremain invisible. Patterns of connectivity can serve as a diagnostic,revealing opportunities to connect those who are disconnected,and ‘spreading’ constructive social norms through highly connectedindividuals whose behaviour is likely to be imitated by those intheir network. 54

The possibility of representing social networks visually also affordsa kind of mirror in which individuals can identify themselves andtheir patterns of social interaction. These ‘sociograms’ serve not onlyto improve objective understanding of how communities function, butalso provide a tool to change the subjective perception of individualsas community members. This generates re exivity that we believemay in itself lead to more pro-social behaviour, better awareness of theconditions in which their actions are taken, and result in a greaterability to shape them. In this sense, viewing communities as socialnetworks is bene cial in terms of both process and outcome. Usingmethods that contribute to creating change alongside interventionsthat are designed on the back of the intelligence they generate is partof the RSA’s wider emerging research identity.

Before we discuss the speci c methods we have used in our researchsites in detail, we offer a thumbnail portrait of each area to assistthe reader in understanding the context within which our research hasbeen undertaken and in which our plans are unfolding.

47 Christakis N and Fowler J, C c : th A az p s c a n a H h sha o l v, 2009,

p32, Harper Collins, italics added.

48 These methodological positions stem from the philosophyof social science rather than political theory, and shouldnot be confused with, for instance, individualism orcommunitarianism. The issue at stake is not ontological(what there is) or ideological (what is most important),but methodological, i.e. which units of analysis lendthemselves to the richest explanations of socialphenomena.

49 Giddens A, th C s c : o hth s c a, 1984, Polity Press.

50 Ho D, Interpersonal Relationships and RelationshipDominance: An Analysis Based on MethodologicalRelationalism, 1998, p1, A a J a s c a p ch .

51 Robins G and Kashima Y, Social Psychology and SocialNetworks: Individuals and Social Systems, 2008, p11, A aJ a s c a p ch .

52 For an excellent overview, see Marin and Wellman, SocialNetwork Analysis: an Introduction, in Ha s c a n A a , Carrington P and Scott J (eds)forthcoming 2010, Sage.

53 Eagle N, Macy M, Claxton R, Network Diversity and

Economic Development, 21 May 2010, sc c

54 Christakis N and Fowler J, Jan 25 2010, op cit.




new Cross gAte

n C ga ca h h hl b h l ha . i hp c ha h ( h lb h s h a ), h a t a hH h h a h c h ok r a a m a f a C h

h. d , l ha C ag h u v a h a .i h ca a cha ,

c n C ga a a t aa a a , a ha h A2, h

aj a l h h a , h h . i a v

a a (soA ) ( a h ) a ha a a

a 9,000 .

th s v soA ha a c a ha h c c ch a aa v a , c c a

a ch . w a soA a a va a -c a a a a h ca

h a c a h , c hh . i h a a a

h a a a a a a. kt a soA c a a h a cha a v v c a h .i h a a h n Cga , h A2 a a a

. Ha cha soA c a a ac va a a ( a hw h C a Ha a h ) ha c a h a cc /

va acc a acq h h . m soA c a a h

a ch ( a ch aHa a h ’ A ’ ).

ov a , v ha n C ga v c a acc a (c a 19% a a ) a 29% a

cc (c a 69% a a ). tha a h ca v : 48% a wh

h c (c a 91% a a ); 13%a b ac /b ac b h Ca a ; a 20% ab ac /b ac b h A ca .

th a a ha c n d a C (ndC) c 2001, h

h a c ma ch 2011. thn C ga t h cc v h c

h ndC a ha a ca c /a

a h v h c a h a a h ndC a .

th ndC a a a j ca a a c a c

a ; ca a a a a ; h a h a c

a c a c h . th a a c h 2001 c

va (imd) a v

ca a a a a ca c a c a , h h v a . d ch ,

, acc h ch c v ht , a c h v a a h ch

ac hach h .

knowle west

k w c v a a h f a s h b .A h h h h

a c b , b a a ab , k w ‘ a h ca

a a ’.55 i c a ca a , kw ca c a h h c , h a a 95%

h c a h a ca h k w , a

v a a a h a a a, h h c a ha ‘ h

h h a , a ha ’.56 l ca , k w a

a ‘w ’, a h a a a a h a a a

c h a a, ca h acha ‘a w ’ ca h a a

h a a j a a .57 d c v a k w c

h va c ac , a h h h h a ,

a v h h ch ch a a c , a aa a c a j c , k wm a C .

Acc ,‘the KnowleWest Media Centre (KWMC) works with thecommunity to develop the creative,educational and social potential of peoplewithin the surrounding area. We producehigh-quality lm, design and media work,provide exciting experiences for young peopleand run a diverse programme of arts activities,including exhibitions, events, talks and screenings. We also provide access to project

facilities, training and mentoring, as well asmanaged workspace for local businesses. ’58



19

3.3 metHods

‘Social network’ makes people think of Facebook and similar forms of social media, but we believe there is still much to learn about of ine,face-to-face connections, and these are the focus of this report. TheNew Cross Gate NDC programme employed mainly of ine methods;whereas the Knowle West Media Centre employs digital methods toa large extent. We have therefore concentrated on our ndings fromNew Cross Gate and offer a fuller account of our methods in thislocation. While we draw on our ndings and analysis from KnowleWest in this report to some extent, a forthcoming RSA paper on thelinks between access to digital information and communication andsocial capital provides more detail. 59

Given the complex nature of social networks, we employed a mixedmethods approach to mapping and understanding social connections.Social network analysis on the scale we have undertaken yields a vastamount of complex data. This report contains samples of our analysisto illustrate key themes which have emerged and our discussion of thetheory. Rather than provide masses of data throughout, we willproduce in-depth brie ngs following this report that interrogate eachof our main themes in detail.

Method in New Cross Gate

In New Cross Gate we devised a questionnaire with which to surveylocal residents. This was based on our reading of the literatureon social capital, social networks and of measurement techniques forboth social capital and networks. Our questions were designed tounderstand the different ways people were empowered and mademore resilient through their local connections. Broadly speaking, thequestions were designed to elicit connections to people respondentsknew well (strong ties), and to those they may not have known so well(weak ties), and to understand where there are no connections(by omission, no ties). We discuss the nature and relative importanceof these ties in section 4.1 (Law 7: Weak Ties Get You Working) of thisreport. For example, respondents were asked questions such as: ‘whoare the people that you seek out for advice or to discuss mattersimportant to you?’; and ‘who are the people you know who seem to begood at bringing other people together?’ A copy of our questionnaireis given in the Appendix. These questions worked on a ‘name generator’ basis. 60 For allquestions, respondents were asked to give up to six names in responseto each question. In addition, respondents were asked to givea numerical score to indicate how valuable each relationship was tothem, and to say whether the people they named lived in their localarea. Local was self-de ned rather than de ned, for example, byadministrative boundaries, in order to re ect respondents’ networksand not limit them to imposed or arbitrary geographies. Care wastaken to consider question order effects and their potential impact onthe validity and reliability of the data. 61

55 As described by Director of Knowle West Media Centre,Caolyn Hassan.

56 Suzanne Lacy, Artist-in-Residence and leader of Universityof Local Knowledge Project.

57 Ibid

58 http://www.kwmc.org.uk /

59 Davies W, th s c a Va d a n d vC , William Davies, RSA (forthcoming).

60 For example, see Lin N, Building a Network Theory of Social Capital. p3-30 in Ronald Burt(ed) s c a Ca a :th a r a ch, 2001, Aldine de Gruyter.

61 Pustejovsky J and Spillane J, Question-order Effectsin Social Network Name Generators, October 2009s c a n , Volume 31, Issue 4.

reseArCH metHods: unCoVering soCiAl networks

http://www.kwmc.org.uk/

http://www.kwmc.org.uk/




Respondents were then asked a series of ‘resource generator’questions designed to understand the availability of social resources torespondents. These included, for example, understanding whetherrespondents knew people who worked at the local council, who coulduse the internet, or could sometimes hire people. Resource generatorsare diªcult to construct as question selection is based on theory andevidence concerning the relative usefulness of particular social assetsto respondents in particular environments. Here, we have employedour overarching framework of empowerment and resilience, informedby academic studies, to create our question selection. 62

Finally, respondents were asked a series of questions about themselves,including age, gender, ethnicity, economic status, household statusand length of residency in the area. This enabled us to comparenetworks and access to social resources for different socio-economicgroups and also to compare the pro le of our sample against theknown pro le of the New Cross Gate population. Our survey was piloted in 2009, and our nal survey was undertakenin late 2009 and early 2010. The survey was conducted by a teamcomprised of RSA staff, postgraduate students from LSE andGoldsmiths University, which is located on the periphery of the NewCross Gate area, and local residents from New Cross Gate. In trainingand brie ng researchers, we attempted to build links (emergingbridging capital between local residents, local students and the RSA)that would carry through to further stages of our work in New CrossGate to co-design and test social network based approaches toaddressing local problems. All research participants were briefed andgiven debrie ng information. Our survey was administered primarily in the super output areas thatmake up New Cross Gate as de ned by the NDC programme that hasbeen running in the area and that is a partner in our work, but also inthe surrounding neighbourhoods into which social networks of NDCresidents extend. Interviews were conducted mainly door to door, withadditional intercept interviews in public spaces. Each set of responses represents a rich ‘ego-centric’ network of connections through which each respondent’s empowerment andresilience can be understood. These individual networks were thenpieced together to build an indicative collective, or ‘global’, network.Some 280 individuals were interviewed, which generated a totalof over 1,400 local people, institutions and places. In addition, thedisposition of being neighbourly, or valuing neighbourliness, isincluded in the 1,400 nodes on our network map.

This mixed-node approach allowed us to explore the relativeconnectedness and centrality of private, public and social assets andto understand how they themselves are networked. The data from thepilot and main survey was integrated and cleaned, with names beingcross-referenced across the dataset to ensure consistent and accurateidenti cation of all nodes that were named by multiple respondents.Networks were constructed and analysed using UCINET andNetDraw software.

This mixed-node approachallowed us to explore the relativeconnectedness and centrality of private, public and social assets

and to understand how theythemselves are networked



21

In addition to our network analysis, in-depth interviews wereconducted with ‘key hubs’, as identi ed by community workers andresidents and by our aggregated social network map. 63 By key hubs,we mean those in the network who are highly connected and who actas bridges between different parts of the network. The interviewswere undertaken by RSA staff and Fellows and served to helpunderstand how they had arrived at these network positions, and theirroles, behaviours and connections in more detail. Interviews wereanalysed thematically.

Finally, we explored the social networks of local residents who hadengaged with and been supported by the Kinship Project, a localinitiative funded by New Cross Gate NDC and the Child andAdolescent Mental Health Services (CAMHS) team at the SouthLondon and Maudsley NHS Foundation Trust. The project works withchildren, young people and parents, directly supporting those who aresocially excluded and/or suffer from low mental wellbeing. It doesso by building networked capacity in the local community to recogniseand support mental health recovery, and building networks to betterintegrate those working with the project into community life andsupport. We explored changes in networks over time of three familiessupported by the project, on a case study basis. In addition, weinterviewed project staff and those in the community who had beensupported by the project in developing skills and knowledge. We alsoanalysed before and after measures of various capabilities, andexplored how networks had supported their acquisition and enabledand sustained their use. This report touches on the ndings from thecase studies where relevant. A full account of this research will begiven in a forthcoming RSA paper to launch a new programme of work to explore and use social networks to address social inclusionand mental wellbeing in various research sites across the country. 64

Method in Knowle West

Our approach in Knowle West was more qualitative. Here, weexpanded our exploration of networks that empower and support toinclude to digital (online) networks and their relationships with social(real world) networks, and the extent to which they are mutuallyreinforcing. Our approach was shaped in part by the resources availablefor our work, which were more limited than in New Cross Gate. We worked with the Knowle West Media Centre (KWMC) which isa key hub in the community and uses digital means to fosterawareness of and engagement with social issues in the area. Ourapproach was two-fold. Firstly, we undertook in-depth interviews withkey stakeholders in the KWMC and local residents who were identi edby KWMC staff and local residents as being individual ‘hubs’ in thecommunity. These interviews focused on local strategic approachesto network building; how networks are constructed and cascadeinformation across the various platforms, relationships and channelsthat make up the local media ecology; and how social and digitalnetworks and communications interact. We were interested inunderstanding local stories as case studies of chains of events that wouldillustrate networks, communication and interaction. Ten in-depthinterviews were undertaken by RSA staff using a semi-structuredinterview guide and were analysed thematically.

62 Van Der Gaag M and Snijders T, th r c g a :

s c a ca a q a ca h c c , 2004,http://www.xs4all.nl/~gaag/work/RG_paper.pdf

63 Our key hubs were an NDC Programme Manager,a Community Representative from the New Cross GateTrust, Green Shoots project manager, Credit Unionmanager, 170 Community Project director, 170 Worksmanager, Barnes Wallis Centre manager, Hat eld TrainingCentre manager, Building Healthier Communities manager,and Kender Tenants & Residents Association (TRA),Somerville TRA, Winslade TRA representatives.

64 In partnership with UCLAN and LSE, the RSA will be

launching a longitudinal study on a social networksapproach to social inclusion and mental wellbeing later in2010. The work is supported by the Big Lottery Researchprogramme and will operate in multiple sites across England.


http://www.xs4all.nl/~gaag/work/RG_paper.pdf

http://www.xs4all.nl/~gaag/work/RG_paper.pdf




Secondly, a survey was undertaken of 100 users of the KWMC in early2010. This was self-administered and collected at the KWMCreception. The survey sought to understand the role of social anddigital networks in how and why residents had engaged with this hub,and how information, action and opportunity had spread throughnetworks. Demographic information on respondents was alsocollected and survey data was analysed using Statistical Package forthe Social Sciences (SPSS) software.

3.4 reseArCH ConsiderAtions We are trying to develop methods that are as researcher- andrespondent-friendly as possible, so that the process of measuringnetworks becomes also one of building social capital. We have soughtto employ robust research methods and rigorous analysis, but thereare several challenges in undertaking research on community-basednetworks, which we acknowledge.

Social network analysis is a growing eld, but attempts to measuresocial networks at a community level are relatively nascent, mainlydue to the methodological challenges involved in studyingcommunities as networks. 65 In his analysis of community networks,Clark highlights a ‘wariness about the pragmatism behind somedecisions about how to capture social networks in the research.’ 66 The main caveat in using social network analysis at a community levelis that the research tool which yields information about socialnetworks is designed to give results about individual networks frombounded populations, allowing researchers to measure the relativestrength and transitivity of connections rather than merely recordtheir existence and indicators of their strength. Such a tool is ideal forthe analysis of organisations, but problematic at a community level. 67

While neighbourhoods can be demarcated at a geographic level,measuring all networks that impact on a community wouldincorporate a potentially huge number of people. As we discuss laterin this report, as a rule of thumb, in uence extends through networkswith up to three degrees of separation; in other words, an individual’sfriends’ friends’ friends have a discernible impact upon them.However, the person at this nal level of remove is in uenced throughtheir social networks in the same way, and so on. Such chains of in uence mean that if we are interested in identifying social assetsand in uences available to a community through its social networks,we may need to survey a huge and unknown number of people tounderstand, for example, the social network reach of an area the sizeof New Cross Gate. In addition, current research methods formapping of ine social networks are labour-intensive and requirerespondents to describe their personal relationships, which can leadto high levels of refusal if the purpose of the study is not suªcientlyclear and aligned in some way to participant interests.

In aggregating the ego-centric networks we collected, we present ourresearch ndings as indicative of the patterns of relationships thatexist, and use this as an evidence base upon which to proceed. We do

not claim that our social network maps represent an exhaustive set of all possible connections between all possible nodes.

65 Canadian Sociologist Barry Wellman was a pioneer inthe eld. See Wellman B, The Community Question: TheIntimate Networks of East Yorkers,1979, A ca J a

s c , Vol 84, p5.

66 Clark A, 2004 op. cit. r a l w pa :u a C , May 2007 http://www.reallifemethods.ac.uk/publications/workingpapers/2007-05-rlm-clark.pdf

67 For example, organisational anthropologist Karen Stephensonuses Social Network Analysis to show the relationship

between organisation hierarchies and heterarchies, andthe Work Foundation produced a report in 2003: mas c a n o a a, by Kelly Drewery andWilliam Davies.









23

As with much research that engages cold contacts in lengthy surveys,respondents experienced an element of survey fatigue during theinterview. Our experience of piloting the survey led us to cut down thenumber of questions asked, and to phase the questions so that allname generator questions came rst, followed by resource generatorquestions, as the latter were deemed to be less important. Analysis of the data reveals that only three quarters of respondents answered allresource generator questions and so we have not included resourcescores in the analysis of the network. Summary

• Our research project has been designed as a response tocommunity interventions that fail to understand or take accountof social networks.

• We have undertaken research into networks in New Cross Gate,a multiply-deprived neighbourhood in southeast London that iscoming to the end of a New Deal for Communities programme;and Knowle West, an isolated, deprived residential area severalmiles outside of the centre of Bristol. Such areas house tensions,problems and hidden assets common to many communities.

• Our approach is guided by methodological relationism;relationships are the principle unit of analysis.

• Social network analysis allows us to focus on patterns of relationshipsand exclusions. It enables us to make visible connections andin uence in a network, and to re ect on our own position and rolein a community network.

• We undertook door-to-door surveys in New Cross Gate to understandlocal social networks, together with in-depth interviews of key hubsin the network. We constructed a network map of some 1,400 nodes(local people and institutions) as an indicative blueprint for how thecommunity works. In Knowle West, we interviewed local keyconnectors and in uencers and surveyed users of the Knowle WestMedia Centre.

• Undertaking social network analysis in communities – large,unbounded populations – is methodologically and practicallychallenging, and yields large amounts of rich, complex data.However, through this project, we are learning how social networkresearch is best done in neighbourhoods.





seCtion 4: soCiAl networks:struCture, funCtion And proCess

This chapter outlines the theory of social network structure andfunction, and examines each aspect of this theory in light of theemerging data from our eldwork in New Cross Gate and KnowleWest. It also discusses the potential bene t of social network analysisand re ection as a process which itself strengthens networks.

The strange yet familiar world of social networks

Physicist and Philosopher n b hsaid that if you were not shockedby quantum theory then you didn’t really understand it. This pointmay be at least partly true for the role of social networks in our lives.We have evolved as social beings, so at a deep level we may be moreor less intuitively adept at managing our social connectivity in waysthat enrich our lives. But we struggle to comprehend our profoundsocial interdependence and to accept its diffuse in uence on ourbehaviour. We are more ready to place ourselves at the centre, to seeourselves as the authors of our own life-stories, rather than accept thedegree to which they may be scripted by people we will never know.

The limitations of our ego-centric psychology and the growingevidence of the power of network effects underpin our argument thatapplying a networked approach to community problems couldstrengthen area-based initiatives. It also informs our hypothesis thatincreasing self-awareness about social networks has a transformativepotential at both an individual and community level. This logic of re exive social awareness also underpins the approach and applicationof the RSA’s Social Brain project. 70

Visualising social connectivity through social network maps could beparticularly powerful. As a species we have evolved with an advancedability to identify patterns in visual data and visual representations arehelpful in making sense of abstract, complex phenomena. Likewiseit is hard to grasp the meaning of emergent phenomena withoutrecourse to structure. But at the same time we must also guard againstthe instinct to infer and impose structure where there may be none,or assume a predictable causal sequence where the pattern is onlydiscernible in retrospect.

So which, if any, patterns can we discern in social networks? Howare their effects felt in real-life communities? And how can we makestrange phenomena more intelligible to our order-seeking minds?This chapter attempts to answer these questions in parallel.We examine each aspect of the structure (connectivity), function(contagion) and process (re exivity) of social networks both in theoryand in the lived reality of the speci c community, New Cross Gate,where we have undertaken primary research.

The data from our eldwork is only partial and indicative at this stage,and this report only a rst step in the process of uncovering theinsights from social network analysis in a non-institutional setting.We will be undertaking and publishing deeper analysis over thecoming months.

I am bound to everyone on thisplanet by a trail of six people.— J h g a , s d s a a68

It transpires that your colleague’shusband’s sister can make you fat,even if you don’t know her.— n ch a Ch a69



soCiAl networks: struCture, funCtion And proCess 25

68 Guare J, s d s a a, 1990.

69 http://www.thersa.org/events/audio-and-past-events/2010/connected-the-amazing-power-of-social-networks-and-how-they-shape-our-lives

70 http://www.thersa.org/projects/social-brain

71 Milgram S, The Small World Problem, p ch t a,1967, Vol. 2, p60-67. Some have suggested that Hungarianwriter Frigyes Karinthy was the original proponent of ‘six degrees’. He outlined the idea in a short story called‘Chains’ published in a collection titled ev h d in 1929.

72 Watts D, Dodds P and Newman M, Identity and Searchin Social Networks, sc c, 2002, Vol. 296; Dodds P,Muhamad R, Watts D, An Experimental Study of Search in

Global Social Networks, sc c, 2003, Vol. 301

73 Christakis N and Fowler J, 2009, op.cit. The authorsoffer three possible explanations for the existence of thethree-degree rule, all of which they feel may be relevant.1) Intrinsic decay: a decay in the delity of informationonce it is transmitted. 2) Network instability: for a personmore than three degrees removed, the chance of one of the three degrees of connection being severed is quitehigh. 3) Evolutionary purpose: we are built for small groupbehaviour, and have not evolved to be connected to peoplefour degrees removed from us. Speaking at the RSA,Christakis also clari ed that the three degree rule should beunderstood as ‘a rule of thumb rather than a social fact.’

74 http://www.thersa.org/events/vision/vision-videos/robin-dunbar-how-many-friends-does-one-person-need

4.1 soCiAl network struCture:lAws of ConneCtiVity

How do social networks operate? What is their underlying structure?Each section below describes a particular facet of social networks asrevealed by the emerging research literature. We seek to take thisanalysis further by assessing each of these attributes in turn for theircongruence with some early ndings from our New Cross Gate studyor for their relevance to community regeneration in general. Thisconnection between theory and practice helps us evaluate howa networked approach can provide new insights into the problemsand opportunities for local communities.

We describe each of the structural attributes of social networks asa ‘law’ of connectivity to aid understanding. However, given theemergent nature of network phenomena, a more accurate descriptionwould be that of a pattern.

la 1: Six Degrees of Separation, Three Degrees of Influ enceIn theory

Stanley Milgram’s 1960’s letter-forwarding experiment 71 in which hesuggested that everybody in the world is connected up to six degrees of separation was considered by many to be unreliable research, but thisremarkable nding was reinforced by a 2002 study using email dataof 48,000 people. 72

It is important to recognise that connection does not mean in uence.We may be connected to each other through six degrees, but

Christakis and Fowler’s research indicates our in uence only extendsto three degrees:

‘Everything we do or say tends to ripple through our network, having animpact on our friends (one degree), our friend’s friends (two degrees) andour friend’s friend’s friends (three degrees). Our in uence graduallydissipates and ceases to have a noticeable effect on people beyond the social

frontier that lies at three degrees of separation. Likewise we are in uencedby friends within three degrees but generally not by those beyond.’ 73

In addition to our six degrees of separation and three degrees of in uence, anthropologist r d a argues that we have the

social and cognitive capacity to have up to 150 friends, the so-calledDunbar number. 74

Taken together, these three points suggest a high degree of interdependence. Hypothetically, a person with 150 friends who hasthree degrees of in uence could in uence and be in uenced by up to1503 (i.e. well over three million) people. In reality, there would beconsiderable overlap between the friends, and in uence decreaseswith each degree of separation, but the point serves to illustrate howthinking in network terms expands our idea of where our in uencebegins and ends.

http://www.thersa.org/events/audio-and-past-events/2010/connected-the-amazing-power-of-social-networks-and-how-they-shape-our-lives



http://www.thersa.org/projects/social-brain

http://www.thersa.org/events/vision/vision-videos/robin-dunbar-how-many-friends-does-one-person-need











The three degrees of in uence rule is one of the main networkheuristics to emerge from a series of in uential papers on a range of topics including obesity, happiness and smoking, in which Christakisand Fowler highlight the extraordinary power of social networks toin uence behaviour, spread values and shape attitudes. 75 Their hugelyin uential body of research arose from the fortuitous discovery of detailed time series public health data in Framingham, USA, whichcontained information on a core group of over 5,000 people withinformation on friends, relatives, neighbours and co-workers, withina larger network of over 12,000 people, going back 30 years andleading to a map of over 50,000 social ties.

In the community

Capturing relational data of this kind, and understanding the powerof in uence, helps us identify the opportunities and limitations fordevelopment in a place like New Cross Gate. Our data there indicatesthat there is generally about two percent of people who do notknow anyone with regards to any aspect of empowerment andresilience covered by our questionnaire (see Appendix), let alone onquestions that explored people’s networks through which they canin uence local decisions. Mapping the disenfranchised in this waycould ultimately help increase social inclusion via better targetedcommunications and policy interventions.

In addition to having their in uence weakened through lack of informalsocial connections, many people are cut off from direct access to formalsources of in uence. For instance, two thirds of people in New CrossGate do not know anyone who works at the local council, anda third do not know anyone who is in a position to employ people evenon a temporary basis. In uencing the media is also a remotepossibility for many, with 40% of people not knowing anyone whoknows someone at a local newspaper, website, TV or radio station.

There are some important differences in this regard betweendemographic groups. Our analysis in New Cross so far indicates thatyoung people (16-24 year-olds) have below average access to socialresources, and that connectivity increases with age until the 65+group, where we see a dramatic drop-off in connectivity. Those overthe retirement age in most cases have the fewest social resources todraw upon.

In terms of household type, single parent families have fewer socialresource connections than average. This nding is consistent withprevious research that shows fewer parental connections impacts onthe future economic prosperity of children from single parentfamilies. 76 Although the numbers in our analysis are small, it seemshalf of single parent households do not know people who hire someonefrom time to time (an indicator used to understand the usefulnessof peoples’ connections), compared to a third of residents overall.

75 Christakis N and Fowler J, The Spread of Obesity in a LargeSocial Network Over 32 Y ears, 26 July 2007, n e aJ a m c ,Vol 357, p370–379; Christakis N andFowler J, The Collective Dynamics of Smoking in a LargeSocial Network, 22 May 2008, i ,Vol 358, p2249–2258;Fowler J and Christakis N. ‘Dynamic Spread of Happinessin a Large Social Network: Longitudinal Analysis Over 20Years in the Framingham Heart Study’, i , 2007, Vol 357,p 370-9, 3 January 2009.

76 Giddens A, th th wa : th r a s c ad c ac , 1998, Polity Press.

77 Krebs and Holley, b s a a Ch h n b , 2002

http://www.supportingadvancement.com/web_sightings/community_building/community_building.pdf

78 Halpern D, 2005 op. cit.

79 Aldridge S and Halpern D, s c a Ca a : A ca , 2002 http://www.cabinetoªce.gov.uk/media/

cabinetoªce/strategy/assets/socialcapital.pdf

80 Taylor M, Neighbourhood Management and Social Capital,2007, Communities and Local Government ResearchReport 35.

81 Giddens A, 1998, op cit.

82 http://www.cabinetoªce.gov.uk/newsroom/news_releases/2010/100722-citizenservice/national-citizen-service.aspx

83 Butler T, ‘In the city but not of the city? Telegraph H llersand the making of a middle class community’, 2008,i a a J a s c a r a ch m h, Vol 11(2).



http://www.cabinetoffice.gov.uk/media/cabinetoffice/strategy/assets/socialcapital.pdf


http://www.cabinetoffice.gov.uk/newsroom/news_releases/2010/100722-citizenservice/national-citizen-service.aspx












27

la 2: Birds of a Feather Flock Together

In theory

Evidence suggests that we shape our own networks, deciding bothintuitively and consciously how many people we connect to, howdense and well-maintained these connections are and how central weare in the networks we create. As such we tend to create a network inour image, associating with people like ourselves in an act of what thenetwork theorists call homophily.

This phenomenon is described in some common gures of speech.Community consultants Krebs and Holley argue that when networksare left unmanaged by residents, community developers and thepublic sector, ‘birds of a feather ock together’ and that ‘those closeby, form a tie’ . This can be problematic because it leads to smallbonded clusters lacking in diversity and access to information, tiesand resources outside. 77

One as yet unanswered question in the existing literature is how bestto understand the relationship between bonding and bridging capital.Halpern, for instance, has suggested that community health dependsupon having both, and there is some evidence that highly bondedimmigrant communities are actually better at integration and formingbridging connections with other communities. 78 A more generalobservation is that excluded groups tend to have high levels of bonding social capital relative to bridging social capital, which is whyyear two of our project will make building bridging capital a priority. 79

Such an approach might mean ‘reconnecting the rich’ as ma ta argues 80, which means not working only within neighbourhoods,but also connecting the af uent with the less af uent and vice versa,both to combat stereotypes and expand access to opportunities.A h g writes about the elective separation of the better off,for instance through gated communities, and warns that such anattempt would not be easy. 81 In announcing plans for the NationalCitizen Service, the coalition government explicitly aims to mixparticipants from different backgrounds to create diverse but cohesivegroups, which alongside delivering social action projects could fostervaluable bridging capital. 82

In the community

Our research reinforces Giddens’ latter point. Part of our eldworkinterviews were also conducted within the more af uent TelegraphHill which adjoins New Cross Gate. The data highlights the practicalchallenges of building bridging capital in this way, echoing the ndingsof a previous Telegraph Hill study conducted by Butler in 2008:

‘I don’t really think there is a chance to get people from Telegraph Hill tohelp. There’s no real desire to bring any of those skills down from the Hillto work here, unless they’re handsomely paid for it. There’s not a lot of give

for nothing, outside of their community.’ 83

soCiAl networks: struCture, funCtion And proCess

...we tend to create a network inour image, associating with peoplelike ourselves in an act of what thenetwork theorists call homophily.




Some of the challenge lies in creatingeveryday opportunities for interaction, whichis part geographical and part habitual:‘You could draw a huge black line betweenTelegraph Hill and New Cross Gate. Not many from Telegraph Hill pass down this way.’

That said, several people living in TelegraphHill who were involved in the study clearlyrecognised the importance of constructivenetworks between the areas. This perceivedseparation was in part attributed by a residentto the way the boundaries are drawn for theNDC area, focusing exclusively ona lower-income part of the London Boroughof Lewisham. Administrative boundaries of wards and super output areas do not meshwith people’s networks, and serve to con nearea-based initiatives in ways that segregateareas and produce strong internal identitiesand bonded capital rather than forging linksto other areas that can foster empowermentand raise aspirations.

In social capital terms, Knowle West is considered a highly bondedcommunity. From neighbourhood statistics we know that 95% areWhite British, almost a fth of 16-74 year olds are in what the Oªcefor National Statistics terms routine occupations, (compared to 9%nationally), and 6% have never worked (3% nationally). 84 These statisticsindicate less opportunity for contact across demographic groups andless in uence from ‘outsiders’.

Until a few years ago there was, as one respondent put it, an implicitassumption that ‘those who grew up on the estate, stayed on the estate’ 85,which is still true, but to a lesser extent. Locally Knowle West residentsare known as ‘Westers’, and there appears to be a disparity between theinternal and external perceptions of the area, encapsulated by the factthat being a Wester can be both a badge of honour anda judgemental label.

In contrast, New Cross Gate has greater ethnic diversity and isrelatively integrated. From our network maps, we observe only minorclustering among people of different ethnicities, which may be theresult of families and particular culturally speci c or faith-basedactivities. This matches the NDC's view that the area is wellintegrated. For example, there were very low levels of racially-motivated hate crime in the recorded crime data for the NDC. 86

In terms of making use of hidden wealth and encouraging behaviourchange, Knowle West’s homophily (when connections between similarpeople are more common than among dissimilar people) poses aninteresting challenge, but also an opportunity. By combining socialnetwork analysis with demographic data, we can begin to see patternsof homophily. But within these clusters, we can also identify bridgesto other clusters, and then attempt to use those bridges to connectwhole networks.

84 http://www.neighbourhood.statistics.gov.uk/dissemination

85 Suzanne Lacy, Artist-in-Residence and leader of theUniversity of Local Knowledge Project.

86 ndC A a v da a, 2008, Social DisadvantageResearch Centre, Oxford University.

f 1: wha a n C ga ?

http://www.neighbourhood.statistics.gov.uk/dissemination

http://www.neighbourhood.statistics.gov.uk/dissemination



f 2: t a h n C ga

29

In our research, we wanted to look at those nodes that already linkresidents from each of the SOAs that make up New Cross Gate. Thisapproach drew out six linking nodes from the data that bridge togetherall areas, namely: ‘helping or being helped by neighbours,’ ‘shopkeeper,’‘New Cross Gate Post,’ ‘New Cross Food Co-op,’ ‘Lewisham College’and ‘Postman’ (see gure one) .

Of these, the generic public gures (shopkeeper and postman) appearto be the strongest bridges between all areas. This suggests thatthere may be some value in pursuing interventions that build uponthe connectivity of such public gures to contribute to communityregeneration activities in more explicit ways. Of the linking nodes,Lewisham College is perhaps the least surprising appearance, beinga well-established further education institution in the borough.

The New Cross Gate Post (a local newspaper) and the New CrossFood Co-op are relatively recent initiatives that have been deliveredand/or supported by the NDC. Perhaps as expected, the NDC itself isnot cited as a mutually recognised part of the networks of residentsfrom each of these areas, although sub-projects of the NDC are.With respect to the New Cross Gate Post example in particular, thisfact suggests that if we value bridging capital in our communitiesthen the NDC was right in its decision to extend the distribution of the newspaper by local residents beyond the formal boundaries of the NDC area and to include neighbouring areas such as Musgrove(see gure two, below) .

la 3: Location, Location, Location

In theory

The third aspect of networks is that they shape us, and have positiveand negative effects depending on our position in the network. Forexample, being central to a network is good for receiving informationand being happy, but bad for avoiding infectious diseases.


...the generic public gures(shopkeeper and postman) appear to be the strongest bridges betweenall areas.




f 3: n C a

At a community level, spreading information about jobs, trainingand recreation are likely to be more salient issues, but in this respectit is important to recognise that in so far as network location matters,geographic intuitions don’t get us very far, as we discovered inNew Cross Gate.

In the community

The geographical map in gure two showsthat the Kender Triangle area is, in spatialterms, right in the centre of New Cross Gate.However the social network map in gurethree reveals that the Kender residents weinterviewed are relatively isolated in networkterms. For example, the green circle in thebottom left of the map highlights a sub-clusterof respondents from Kender Triangle(represented by the darker grey nodes), at theperiphery of the network.

In network terms, these Kender Triangle residentsare placed together, just outside the outercore of the network, which is represented bythe yellow circle. Within the central core – thecoloured nodes within the centre white circle –there is a preponderance of residents from theSomerville area of New Cross Gate(represented by darker green nodes). It would seem that, to an extent,Somerville acts as a bridge between Winslade and Kender Triangleand the wider New Cross Gate area. The location of this bridge issurprising, given that Somerville does not lie geographically betweenthem. As we have noted before, communities and networks oftentranscend physical boundaries.

Overlaying geographic data onto social network data can enhance ourability to target interventions. It can help to identify where and how toreach the intended recipients – often obliquely - through the chains of tiesand in uence that make up the local social structure.

la 4: Imitation Drives Contagion

In theory

This attribute of networks concerns peer-to-peer mimicry. Subconsciouslyor otherwise, we end up aping what our friends do and this mimicryis in turn passed on to others in the chain, in uencing people severaldegrees of separation removed from ourselves.

This may be the basis for pressure to ‘keep up with the Joneses’, withfashion trends and the pressures on children to keep up with the latestfads. Cultural psychologist m cha t a has described imitationas the link between biology and culture, and it is this adaptive traitthat ultimately underpins the importance of social networks for policymakers. We seem instinctively to want to copy other people and

we are most likely to copy the behaviours of people we see and hearevery day.87

87 Tomasello M, th C a o H a C ,1999, Harvard University Press.

88 Cacioppo J and Patrick W, l : H a na ah n s c a C c, 2008, WW Norton and Co.

http://www.thersa.org/events/vision/vision-videos/professor-john-cacioppo---connected-minds-loneliness,-social-brains-and-the-need-for-community

89 t p ac c, Demos, June 2010.

90 Kahneman D, 2010, op. cit.

91 DiMaggio P, and Powell W, ‘The Iron Cage Revisited:Institutional Isomorphism and Collective Rationality inorganizational Fields’, A ca s c ca r v 1983 Vol 48.

A a h a 2 –- v a

a h

c ( a )

k :k t as vt a h Hw aHa chaC wa

mm v

b c hm c wao hn v









31

In this respect, social network theory is complemented by theemerging eld of social neuroscience, which is helping us make senseof why social connectivity should matter so much at a cognitive andemotional level. As social neuroscientist J h Cac , who recentlyspoke at the RSA explains:

‘The telereceptors of the human brain have provided wireless broadband

interconnectivity to humans for millennia. Just as computers have capacitiesand processes that are transduced through but extend beyond the hardwareof a single computer, the human brain has evolved to promote social andcultural capacities and processes that extend far beyond a solitary brain.To understand the full capacity of humans, one needs to appreciate not onlythe memory and computational power of the brain but its capacity for representing, understanding, and connecting with other individuals. Thatis, one needs to recognise that we have evolved a powerful, meaning-making social brain.’ 88

The fact that imitation is central to network effects was indirectlyunderlined by a recent study by Demos which indicated that trust wasnot ‘networked’; for example, one person trusting a local councillorhad no impact on whether his friends and family trusted thecouncillor in question. 89 This might seem to indicate that socialnetworks are not so important after all, but in fact the Demos researchsuggests that network contagion may depend upon literally seeing(and to a lesser extent hearing) other people becoming fatter, smokingless, smiling more etc. The reason trust does not spread is probablybecause trust is relatively invisible and cannot be ‘copied’.

Christakis and Fowler, in their experiment to test whether cooperativebehaviour was contagious, found that the punishments for non-cooperation did not seem to in uence participants. It seems that weobserve the fact of cooperation, but not the reason for it. 90 It could bethat observing cooperative behaviour is a powerful and underutilisedform of communication.

Imitation also works at the organisational level. DiMaggio and Powell(1983) argue that along with normative and coercive pressures,organisations also experience mimetic pressure. While normativepressures in uence acceptable organisational behaviour, and coercivepressures are exerted by central and local government (or generally,the more ‘powerful’ partner in the relationship), mimetic pressuresmean that organisations imitate others for acceptance, status and

prestige.91

DiMaggio and Powell argued that these pressures result inincreasing organisational isomorphism over time. In other words,organisations experiencing similar pressures and collaborating would‘converge’, adopting similar language, practices, interests andbehaviours over time.


...social network theory iscomplemented by the emerging

eld of social neuroscience, whichis helping us make sense of why

social connectivity should matter so much at a cognitive and emotional level.




The nature of networks is often in uenced by the practices andstructures of community organisations. Small (2009) found that thenetworks of people who engage with community-based organisationsare dependent on the practices and regulations of those organisations,including seemingly trivial protocols. In his study of childcare centres,Small found that, for example, the way meetings were held, how often

eld trips were undertaken, the formality of language, the timesat which children were picked up and dropped off, and the extent towhich users of the centre were able and encouraged to loiter, allin uenced the availability of social capital and were often mechanismsfor producing social inequality. 92 Thus through imitation, there isa danger of perpetuating ‘unhelpful’ practices, particularly if theirpotential impact is not realised.

In the community

Our work to date focuses more on social networks than individualpsychology, although we plan to integrate our work with theRSA Social Brain project later in the programme. Consequently, oursocial network data has not tracked contagion, but there are indicativeexamples from two strands of our research.

Imitation and social learning was evident at the organisationallevel within local community and voluntary groups. As the head of one community organisation from New Cross Gate commented:

I suppose I tried to re-fashion the way our organisation worked on (one of thecommunity groups in the area) who had been successful in winning public contracts and getting funding from the NDC. It gave us a way to develop andto get a seat at the table. It worked to an extent, but we had to start using the‘right words’ and talk about outputs and budgets and so on. It made us more

professional, but I worried that we were losing a bit of our identity and we hadto be careful about losing the things that made us attractive to the communityin the rst place.

In this way, ABIs and the public sector can foster a mutually reinforcingsystem of mimetic, normative, and coercive pressures throughfunding requirements, processes and language. Imitation can drivethe professionalisation of organisations within this system,but can also foster behaviours and practices that serve to exclude.Imitation is a powerful mechanism for the contagious spread of socialphenomena that needs to be understood in the design of interventions,particularly those concerned with behaviour change and the

development of the Big Society. Such interventions need to be visibleand amenable to mimicry. Our recent report into tackling anti-socialbehaviour, ‘The Woolwich Model’, proposes that observing andexperiencing people acting in mediating con ict helps such behaviourto spread. In this sense, actions speak louder than words andinterventions should focus more on spreading behaviour by exampleand not simply through rhetoric. 93

...in so far as network locationmatters, geographic intuitionsdon’t get us very far, as wediscovered in New Cross Gate



33

la 5: It’s Not What You Know, it’s Who They Know

In theory

This network law tells us that our friend’s friend’s friends affect us,in the same manner that our friend affects us, but to a differentdegree, in what is called hyperdyadic spread, a phenomenon that hasjust recently been possible to observe in large social networks.

We have learnt in the previous law of imitation that we are in uencedby indirect ties because imitation is a contagious mechanism.Although I may not see or know people who are my indirect ties, it isenough for somebody I do see to have seen them for me to ‘catch’some aspects of whatever behaviour they are propagating. While weused to say that ‘it’s not what you know but who you know’ thatmatters in life, we now recognise that this is true because of who theyknow, and how well.

Reed’s Law captures the power of hyperdyadic spread, and says, toparaphrase the technical language, that network growth is geometric(2,4,8,16,32, etc.) rather than arithmetic (1,2,3,4,5, etc.). Networks arecomprised of groups of varying sizes rather than individuals, soconnecting with one group invariably means connecting with more thanthat group alone, if only because each individual member of a grouptypically belongs to other groups too. Membership of one group breedsmembership of another in a multiplier effect that expands the growthand strength of the network exponentially. This is a noteworthy pointgiven that l w , government advisor on the Big Society, has statedthat his principal goal is for everybody in the UK to be part of somesort of group.

The importance of this idea becomes clearer if we consider the shapeand nature of networks. Broadcast networks are the most basic form,amounting to a ‘one to many’ network in which my network iseverybody I know, and spreading the word across my network merelymeans broadcasting information rather than an exchange of information. Transaction networks, or one-to-one networks, are morecomplex, featuring an exchange of information between two people.However, the most powerful form of network is the many-to-manynetwork, also known as group forming networks or GFNs. Thesenetworks are at the heart of Reed’s Law, because it is the connection toanother group that signi cantly increases your resources. Thisbridging or linking capital is the kind of connectivity from which mostpeople bene t. 94

In the community

To give some geographic context for Reed’s Law and the idea of hyperdyadic spread, the following diagram (see gure four overleaf) illustrates the myriad of connections between a variety of communitygroups, people and services in New Cross Gate. Our exploratoryresearch suggests that any contagion will not spread in predicableways, because network analysis enables us to see disconnects ina community that would otherwise be invisible.

92 Small M L, u a c a ga – q a v a , 2009, Oxford University Press.

93 http://www.thersa.org/__data/assets/pdf_le/0004/327082/0882RSA_21CE_benrogers_web.pdf

94 Reed expresses his own law in the following technicallanguage: ‘Let’s say you have a GFN (Group FormingNetwork) with n members. If you add up all the potentialtwo-person groups, three-person groups, and so on thatthose members could form, the number of possible groupsequals 2n. So the value of a GFN increases exponentially,

in proportion to 2n. I call that Reed’s Law. And itsimplications are profound.’ ‘The Law of the Pack’, Ha vab r v Feb 2001 http://projects.rsablogs.org.uk/index.php?s=reed%27s+law


http://www.thersa.org/__data/assets/pdf_file/0004/327082/0882RSA_21CE_benrogers_web.pdf


http://projects.rsablogs.org.uk/index.php?s=reed%27s+law









For instance, the two key providers of community services locally, the NDC and the170 Community Project, both appear in theless connected left-hand column. Therelatively small number of connections tothese two bodies is striking. Within this it isworth noting that most of these connectionsare to Somerville residents, with theremaining connections to Kender Triangleresidents. Winslade residents included in thisrepresentation do not cite connections toeither of these projects, despite Winsladefalling within the NDC boundary.

One node which we might expect to constitutepart of the Winslade residents network is theMillwall FC Community Scheme; given theproximity of The Den, Millwall’s homeground, to Winslade. But in this representationof the data the only SOA residents to whom thescheme is not connected is Winslade. Here,then, in two instances we see that the spatialrelationships that we might expect (i.e. connections between Winsladeand the NDC and the Millwall scheme, given geographical proximity)do not actually happen in network terms.

The diagram also indicates that two less densely connected nodes towhich Winslade residents do connect beyond the level of connectivitywe might expect are the public gure ‘Janitor/Caretaker/MaintenanceWorker’ and ‘School Board or Parents’ Association’. If a particularemphasis were put on trying to integrate Winslade residents into thewider community, for example, then social network analysis enablesus to identify particular contexts or individuals through whichconnections might be most effectively made.

The hyperdyadic spread could be useful in situations where policymakers wish to change behaviour, but need to nd a mechanism fordoing so that is relatively eªcient. Christakis and Fowler comment:

‘If we wanted to get people to quit smoking, we would not arrange them ina line and get the rst one to quit and tell him to pass it on. Rather, wewould surround a smoker with multiple nonsmokers, perhaps in a squad.’ 95

While this suggestion sounds slightly Orwellian, the point is not somuch about coercion, but rather a broader notion that eªcientlyeffecting behaviour change requires an understanding of the structureof a network, in order to work with individuals whose behaviour ismore likely to in uence other relevant parties, and then call on thepower of social contagion to increase your chances of success.

In our eldwork we were struck by the signi cance, in this digital age,that residents ascribed to ‘tree posters/notice boards’ as a local meansof communication.

f 4: sc H a c s a n C ga



35

It appears that not only was this practice prevalent, as noted by manyof the eldworkers involved in this research, but also that it wasthe medium usually mentioned as signi cant in response to thequestion: ‘Where do you go or who do you speak to in order to nd outwhat’s going on in your local area?’ In this way, such technologicallysimple methods can be used to build bridges into groups forindividual citizens, and for otherwise unconnected groups to join orbuild common members.

la 6: Experimentation Gets Results

In theory

Networks have emergent properties such that the whole cannot beunderstood with reference to the parts, nor can you predict thebehaviour of the whole with respect to the parts, or vice-versa.Networks operate like any other complex system, in which the pathbetween cause and effect cannot be reliably traced.

Recognising communities as networks therefore brings a degree of legislative humility. Policy makers have to accept that in so far ascommunity policy identi es communities as networks rather thanareas, the outcomes of their policies are in principle much lesspredictable.96 This is diªcult for governments to tolerate, but, asnumerous commentators have indicated, the acid test of whethergovernments can really embrace localism is the extent to which theyare willing to let go not only of the traditional processes of policymaking, but also of ownership and certainty in the outcomes.

A lack of predictability is problematic given that governmentdepartments are accountable for the money raised through taxation,and need to justify their spending choices. Rationales based on‘evidence’ tend to be the most credible, and what government ideallywants is therefore evidence that says: ‘X leads to Y; we care about Ytherefore we are justi ed in spending on X.’

In the community Policy-makers do not necessarily expect the world to conform to suchrough logic, but nonetheless departmental structures and budgettimetables lend themselves to a relatively linear model of thinking, inwhich one chooses a metric of success and then strives to see it

increase. This approach may perpetuate problems in policy areaswhere awareness of complexity and emergence is key. As we haveindicated above, this appears to be the case with communities, whichare not areas but systems, not groups of individuals with stabledemographic data, but networks with unstable patterns of connectivity.As community development consultant A g chputs it:

‘One of the hallmarks of an open, complex adaptive system is itsunpredictability…Attempts to monitor the effectiveness of communitydevelopment using prescribed targets and timescales is thereforeinappropriate, although indicators which track changes in capacity,connectivity and cohesion are probably valid.’ 97

95 Christakis N and Fowler J, 2009, op. cit, p131.

96 Ormerod P, 2010, op cit.

97 Gilchrist A, 2009,op. cit. p127


Policy makers have to accept thatin so far as community policyidenti es communities asnetworks rather than areas, the

outcomes of their policies are inprinciple much less predictable




Connect through your similarityand pro t from your diversity.— k a H 100

An example of this is seen in the signi cantly different impacts of similar projects in similar areas. Both New Cross Gate NDC andClapham Park NDC areas ran Timebanking projects to encouragevolunteering, and Lifestyle Opportunities for Older People (LOOP)projects to encourage participation in healthy activities and lifestylechoices among older people. In both cases, project proposals were verysimilar and both had similar resources and implementation plans. InNew Cross Gate, the LOOP project established a series of healthylifestyle and health awareness activities resulting in improved levels of health among older people. The project failed to achieve such impactsin Clapham Park by a considerable measure. 98 Conversely, theClapham Park Timebank project resulted in a signi cant shift in theoverall levels of volunteering in the area, compared to no signi cantshift in New Cross Gate. 99

Such substantially different outcomes make lessons for replicatingand scaling activities, sharing good practice, and evaluation muchmore diªcult. A further reason for promoting the role of socialnetwork analysis at a community level is that it provides a tool to trackemergent properties, and allows for the measurement of the shapeand directionality of change without relying on linear accounts of causality. In this way, more helpful lessons can be gained about theconditions for supporting successful projects and also about thedomino effects of change.

la 7: Weak Ties Get You Working

In theory

Ties between people vary in strength, and the nature of our connectionsis often more important than their number. While the network ‘laws’outlined above help to clarify the explanatory power of social networks,it is probable that the single most important element of network theoryat a community level is what are known as ‘weak ties’.

Mark Granovetter’s seminal work on ‘weak ties’ highlighted thatopportunities, particularly employment opportunities, are more likelyto arise from people we don’t know very well who have connections toother networks. 101 For this reason, A g chsuggests thatcommunity development is mainly concerned with strengthening andextending the ‘weak ties’ or, in social capital terms, the ‘bridges’ and‘links’, particularly in situations where people nd it diªcult to meet

and make connections.102

However, weak ties are strong only when they connect at least twonetworks that have strength, as A C a indicates: ‘Not all weak tiesare important, however, but rather those that act as bridging ties betweentwo different networks of strong ties, along which ideas, innovations,information and artefacts ow.’ 103

Identifying and making use of weak ties appears to be relevant tocommunity regeneration for three main reasons. First, asGranovetter argued and as a recent IPPR report indicates,employment opportunities are more likely to arise from weak ties.



37

This point is particularly relevant in the context of NDC evaluation,which showed very little positive impact on employment. Socialnetwork analysis provides us with the micro-level data to make thiskind of analysis and design of intervention possible. 104

Secondly, weak ties are a key element of social cohesion. Granovetterexplains that ‘Weak ties play a role in effecting social cohesion. When

a man changes jobs, he is not only moving from one network of ties toanother, but also establishing a link between these.’ 105 He adds: ‘The morelocal bridges…in a community and the greater their degree, the more cohesivethe community and the more capable of acting in concert.’ 106

Thirdly, the signi cance of weak ties should inform arguments aboutsocial justice, because we now know that those in poverty, to theirdetriment, rely on strong ties more than other sections of the population:

The heavy concentration of social energy in strong ties has the impactof fragmenting communities of the poor into encapsulated networkswith poor connections between these units; individuals soencapsulated may then lose some of the advantages associated withthe outreach of weak ties. This may be one reason why poverty isself-perpetuating. 107

These three points seem particularly persuasive with respect to ourcontention that de ning community with arbitrary, or even logical,geographic boundaries may perpetuate certain social problems, andalso suggests a profoundly ineªcient use of resources. However, whilerecognising the value of weak ties is critical, they should not beovervalued as ends in themselves. Recent government rhetoric on theneed for ‘strong communities’ is appropriate in this regard, as long asit recognised that community strength does not mean monolithicgeographical areas where everybody knows everybody else, butmultiple and diverse community networks that are strong insofar asthey have points of contact (weak ties) through which they can shareskills, information and support.

In the community

Considering the signi cance of weak ties for employment, inNew Cross Gate we discovered that, whether as cause or consequence,unemployed respondents appeared to have smaller networks, andwere more likely to be on the periphery in network terms (see gure

ve overleaf). Those not in work were more likely to be found far fromthe core of the network, and were less well-integrated. Similarly,retired individuals were also more likely to be less central to thecommunity network, and consequently more likely to be isolated.

We also found that within the core network, those ‘in work’(light blue nodes in gure ve and ‘in education/training’ (dark bluenodes) appeared to be relatively tightly clustered. It is possible thatthis re ects differences in bonding capital in these groups, with thoseeconomically active more likely to associate with each other andgain advantageous links, opportunities, access and information, andfurther ‘weak ties’ as a result of their more central location in

the network, and tendency to cluster with those who have similarlyadvantageous connections.

98 eva a h loop j c, New Cross Gate NDC, 2006

99 eva a h t a p j c , C a ha pa ndC,2006; and NDC Household Survey 2006, 2008, Ipsos-MORI

100 Krebs V and Holley J, ‘b s a Ch h n w av ’, 2006, http://www.orgnet.

com/BuildingNetworks.pdf

101 Granovetter M, Cha J : Cha mi a a s a C , 1970, Doctoraldissertation, Harvard University; Granovetter M, 1973,op cit; Granovetter M, 1983, op cit; Granovetter M, ‘TheImpact of Social Structures on Economic Development’,2004, J a ec c p c v, Vol 19 (1).

102 Gilchrist A, 2009, op cit.

103 Clark A,‘r a l w pa : u aC ’, May 2007 http://www.reallifemethods.ac.uk/publications/workingpapers/2007-05-rlm-clark.pdf

104 Sumption M, 2009, op cit.

105 Granovetter, 1973, op cit.

106 Ibid.

107 Granovetter, 1983, op cit.


http://www.orgnet.com/BuildingNetworks.pdf











Our data suggests that unemployment is correlated with low levels of connectivity, which is not surprising, but we found that potentiallyempowering social networks can be further eroded through high levelsof transient populations that often characterise target neighbourhoods.The proportion of residents who have lived in the same area for lessthan a year, for example, is about 25% higher in New Cross Gate thanthe national average. In our qualitative research, the transience of the local population was felt to be a barrier to building the networks thatmight strengthen the community. As one disgruntled local residentput it: ‘People are ghting to leave the area because of the violence… the people who move in don’t give a hoot about the area, as they’ll only be here for a couple of years… I don’t know my neighbours anymore.’

4.2 understAnding networkfunCtion: ContAgion

‘The phenomenon of ‘social learning’- learning through observation andinteraction with others- occurs widely in various forms… Many of thedecisions we make are based not so much on the independent rationalcalculation of the costs and bene ts of different actions- the mode of behaviour posited in economic theory- but on observing and copying others.” — pa o 108

In theory

Patterns of connectivity have considerable diagnostic value. They alsohave a prescriptive element, because a variety of attitudes and habitsspread through social networks. For example, experimental research byAustrian economist e f hsuggests that most people are ‘conditionalaltruists’ who will cooperate if they believe others will reciprocate. 109 However, populations are diverse and contain their share of ‘egoists’, whotend to collaborate less. These people strongly in uence the likelihood of others cooperating. In a related study, analytical sociologist p H ö has shown that, with just 5% of a population acting as ‘egoists’, social

interaction effects between them and conditional altruists can lower therate of overall cooperation in a network by about 40%. 110

k ( q a ) a c (c c ):

i ca / ai r

o hun v

f 5: n r a e a t a ...the transience of the localpopulation was felt to be a barrier to building the networks thatmight strengthen the community



39

More recently, a controlled experiment by Christakis and Fowler 111 attempted to discover whether witnessing cooperative or uncooperativebehaviour might promote changes in cooperative behaviour evenwhen reputations are unknown and reciprocity is not possible. Theresults suggest that the contagion of behaviour is path dependent,and it appears that mimicry trumps conventional rational self-interested behaviour:

‘A consistent explanation for both the experimental investigations andthe observation studies is that people mimic the behaviour they observe,and this mimicking can cause behaviours to spread from person to person.If anything, it seems likely that people who are willing to copy strangers’ behaviour in a controlled experiment may be even more likely to copybehaviour observed in real-world settings.’ 112

The contagion of behaviours in social networks is important for anygovernment policy based on behaviour change, and delivering ‘morefor less’. Government cannot, for instance, tackle crime and antisocialbehaviour, unemployment, unhealthy life-styles, chronic ill-health,mental illness, inter-ethnic tensions, or reduce carbon emissionswithout the active participation of the public.

The kinds of behaviour change governments need in order to achievepolicy objectives often arise precisely because of ‘contagion’ spreadingthrough social networks. For instance, with respect to public healthobjectives, b a wa has demonstrated that by unconsciouslyimitating people around us all the time we develop habits of unnecessary eating, wherein we eat more than we might want or needsimply because we are surrounded by other people who are over-eating. 113

In terms of improving road safety, Christakis and Fowler suggestthat if you want more people to wear seatbelts, the message shouldbe focused on people near the centre of social networks, so that thebehaviour is more likely to spread. This strategy of using networkinformation to effect social change applies more generally. 114 Inpractical terms, having the blueprint of community networks enablesus to locate and work with key in uencers and meta-networkers, assuggested in the example below.

In the community: Familiar Strangers: Dustmen, LollipopLadies and Quiz Masters

In exploring the core nodes of New Cross Gate, we found that publicgures are central to the social networks here. In particular, public

servants — postman, police oªcer/Police Community Support Oªcerand dustman — rank particularly highly in terms of how connectedthey are. This nding highlights the relative importance of publicservants compared with formal political gures.

The postman (as a generic public gure, not the same postman eachtime) has the highest ‘coreness’ value in the New Cross Gate network,and is potentially a key link in any network-based strategy forcommunity regeneration. Preliminary ndings indicate that, in networkterms at least, these individuals and others like them (for examplelollipop ladies) wield signi cant latent power in the community. Currentcommunity initiatives should therefore consider how they can betterharness such centrality and power to serve community needs.

108 Ormerod P, 2010, p14

109 Fehr E and Fischbacher U, Altruists with Green Beards,2005, A a & k , Vol 27

110 Hedstrom P, ‘Actions and Networks: Sociology that ReallyMatters…to Me’. 2007, s c ca, Vol 1

111 Christakis N and Fowler J, Jan 25 2010, op cit. Theexperiment was based on the Fehr-Gaechter Public GoodsExperiments, in which participants were placed in groupsof four, and each subject was given 20 money units(MUs) and they had to decide how many to keep and howmany to contribute to a group project.

Each MU spent is costly to the individual but bene cialfor the group. If each group member keeps all 20 MUs,they will make a maximum of 20 Mus, but if each membergives away 20 Mus, they will each earn 32 MUs. Despitethe opportunity to improve outcomes, each individual canalways earn more by contributing less.

Subjects made their decisions without initially seeing theirfellow group member’s decisions, but all contributionswere revealed to each group member at the conclusionof the game, along with payoff outcomes. In an alternateversion, subjects played an identical game, but afterviewing their fellow group members’ contributions,and were allowed to spend one MU to ‘punish’ anothermember, lowering their income by 3 MUs.

In both versions, subjects participated in six repetitionsand the repetition showed that contributions tend todecline in the basic public goods game and to increase

in the public good games with punishment. However,crucially, the research design enforced strict anonymitybetween subjects, and no subject was paired with anothersubject more than once.

This lack of repeated contact made it possible todistinguish between punishment, direct reciprocity,indirect reciprocity and costly signalling on the effect of cooperation.

112 Ibid.

113 Wansink B, Mindless E ating: Why We Eat More than we

Think, 2006, Bentam–Dell

114 Christakis N and Fowler J, 2009, op cit.





115 http://www.thersa.org/about-us/media/press-releases/the-big-society-approach-to-anti-social-behaviour

116 http://www.thersa.org/events/vision/vision-videos/nicholas-christakis-connected

117 Gilchrist A, 2009, op cit. p56

118 Delanty G,C (l ), in Clark A, 2007 op cit.

119 Based on the work of RSA research fellow Stephen Feberas part of Yorkshire Forward’s Renaissance Programme,detailed in The Regeneration Game, RSA Journal Spring 2009.

120 Giddens A, ‘b l a r h ’, 1994, Polity Press

121 Grist M,s : ma b hav h hi c , e v a r a, June 2010, RSA

122 Thaler R and Sunstein C, n : i v d ca H a h, w a h a Ha , 2008,Yale University Press

123 p a r a Cha b hav : hs a k a i ca p c

p c, February 2004, Cabinet Of ce Discussion Paper,Prime Minister’s Strategy Unit; m ac : i cb hav th h p c p c, March 2010, CabinetOf ce Discussion Document, Institute for Government.

Such individuals could ultimately become not only service providersin an instrumental way (the dustman sweeps the street) but also inmore indirect ways by communicating local information, acting inpro-social ways that can be mimicked, and reporting local sentimentsback to the local authority. Acquiring this network-consciousnesscould help to change the approach to such public servants, and openup the possibility of greater co-production and ‘more for less’ in termsof delivery.

For example, the RSA report ‘The Woolwich Model’ looks at thepotential for reducing anti-social behaviour by training such frontlinepublic servants (and other ’core’ community members) in the skills of situational analysis, con ict resolution and management andself-protection that the police already acquire and deploy as a matterof course. 115 It also raises the importance of including social skills andnetwork weaving skills in the requirement criteria and the trainingprogrammes for employing such public servants. These ideas arerevisited towards the end of the report.

Network structure and function: a word of cautionand reflection

Although the evidence for social contagion is strong, and growingstronger, it is important to realise that networks in themselves aremorally neutral, and only serve to ‘magnify whatever they areseeded with.’ 116 A related cautionary point is that networks are whatwe make them:

‘Networks are not always the organisational panacea that many envisage,often failing to ful l their intended function. They contain patterns of prejudice, preference and power because they are based largely on personalchoices that are both ‘tactical and strategic’.’ 117

Nonetheless, we believe the evidence on connectivity and contagionprovides a timely opportunity for increasing eªciency. We need torecognise that aspects of network structure, particularly homophilyand weak ties, should inform any initiatives directed at creatingemployment. In terms of hidden wealth, we can use social networksas a more powerful diagnostic tool to identify patterns of isolation andopportunities for care-based exchanges. In terms of behaviour change,we need to recognise the importance of imitation as the meansthrough which positive behaviour change can spread, particularly withregard to health outcomes. In addition to these major ndings whichare now well established, social networks also create an interestingand relatively unexplored opportunity to increase community re exivity.

http://www.thersa.org/about-us/media/press-releases/the-big-society-approach-to-anti-social-behaviour


http://www.thersa.org/events/vision/vision-videos/nicholas-christakis-connected








41

4.3 Community reflexiVity: ne twork AnAlysisAs botH interVention And diAgnostiC

Community today is abstract and lacks visibility and unity…(it is) morean imagined condition than a symbolically shaped reality based on xedreference points.118 — g a d a

Social networks can be visualised. Our own exploratory research hintsthat the process of visualising networks through participatory methods– getting people to map their own community connections, forexample – can in itself increase the strength of connections in localcommunities. 119 As A h g indicates, such re exivity is anincreasingly pervasive feature of modern life:

‘Social re exivity is both condition and outcome of a post-traditionalsociety. Decisions have to be taken on the basis of a more or less continuousre ection on the conditions of one’s action. ‘Re exivity’ here refers to the useof information about the conditions of activity as a means of regularlyreordering and rede ning what that activity is.’ 120

In a post-traditional society we are increasingly aware of the conditionsof activity (i.e. their underlying principles); we don’t just do thingsunthinkingly but have to make choices in light of this awareness.Moreover, our choices de ne what activities mean. If two gay menmarry they are taking the underlying principles of marriage (commitment,

delity) and reinterpreting them, whilst at the same time changing theinstitution of marriage. 121 In a similar fashion, if we look at the socialnetwork of a community we live in, we become more aware of theconditions of our activity, and change the nature of community as wehave come to think of it. For this reason, in year two of our project wewill research the impact of visualising social networks on social re exivity.

The RSA Kinship Project case studies outlined in section 3.3 alsoindicate that visualising a social network representing improved socialinclusion and mental wellbeing in itself reinforced those feelings. Wedo not know how this impact decays or changes behaviour beyond thereported experience, but the momentary re exive practice did producethe outcome that longer term intervention was trying to bring about.

In some research, people who are asked to re ect on the experience

of participating in it later report enhanced outcomes. As Thaler andSunstein have noted, although designed primarily to cataloguebehaviour, surveys actually affect people’s conduct through a ‘mere-measurement effect’. 122 Part of this phenomenon is explained by the factthat when people are asked what they intend to do, they become morelikely to act in accordance with their answers. For instance, by askingpeople about who or what they know, individuals are far more likelyto re ect on their relationships and change their behaviour as a result. 123 As one respondent in New Cross Gate put it:

‘Going around asking questions about networking to people who should bedoing it can prick the conscience of organisations to strengthen and build on

the loose networks we already have.’





Re exivity has important implications for the way that collectedinformation is presented. Indeed, the framework for effectiveparticipation set out by Lowndes et al. indicates feedback as a simplebut highly effective behaviour change tool. 124 The research processof social network analysis therefore has the potential to be sociallybene cial, but so does the research outcome because the resultingnetworks can be visualised and mapped for individuals and thecommunity as a whole to see.

The map provides a mirror through which people see patterns of connectivity and possibilities for contagion. This process of visualisation may in itself make people view themselves in a morepro-social way.

Given that social networks are potentially viewed as private assets asmuch as public goods, 125 this consideration should inform the natureof the sociograms we use to encourage re exivity. Ego networksportray connectivity as a private asset, while global networks, in whichwe see ourselves as part of a larger whole, may be more likely to makeus think of networks as a public good.

Another important point concerns the research ethics of sociograms,and how we respect peoples’ privacy. Many may want to take part insocial network analysis and know about local social networks, but maynot want their own social network to be on public display. We areconsidering how best to address these issues on an ongoing basis, andin our social network research in New Cross Gate we coded the databy number so that people who wanted to see their ego network coulddo so, but individuals are not named in any of our public-facing maps.

Summary

• Social networks are dynamic, complex systems that operate inunpredictable, strange and as yet unclear ways.

• However they appear to be governed by certain structural(connectivity) and functional (contagion) characteristics thatin uence the ow of information and behaviour throughthe network.

• Understanding these principles, and designing interventionson this basis, may increase our chances of deliveringpositive outcomes through a combination of public policy andgrassroots activism.

• The very act of undertaking social network analysis ina community, and presenting the community with a visualrepresentation of their network, may itself be bene cial forencouraging pro-social behaviour.

The map provides a mirror through which people see patternsof connectivity and possibilities for contagion. This process of

visualisation may in itself makepeople view themselves in a morepro-social way.



43

seCtion 5: Applying networkprinCiples to Community problemsIn section four we outlined the structure, function and process of social networks (connectivity, contagion and re exivity) and indicatedwhy this approach seems timely in terms of addressing the challenges

of eªcient resource allocation, making better use of hidden socialwealth and fostering positive behaviour change.

The second year of our programme is geared towards using ournetwork data from year one as a platform to develop community-generated initiatives and policy ideas. At this stage we can highlightthree broad areas of community policy which would bene t froma network-based understanding of the problem, as well as from usingthis approach as part of the solution.

Firstly, we look at the problem of loneliness and social isolation, andthe potential to draw on untapped sources of support. Secondly, we

look at sources of community resilience and adaptability as a formof hidden wealth. Thirdly we examine the potential for networks toenhance community empowerment – the ability of an individualor community to be in charge of their own life/lives and to shape theactions of public bodies – by using the leverage that exists at the coreof the community’s social network.

5.1 supportiVe Communities: tACklingloneliness And isolAtion

‘Evolution fashioned us not only to feel good when connected, but to feelsecure. The vitally important corollary is that evolution shaped us not onlyto feel bad in isolation, but to feel insecure, as in physically threatened.’ — J h Cac a w a pa c126

Combating isolation ought to be viewed as a goal of regeneration.Even though isolation affects a relatively small number of people,it directly places a strain on public resources because, as Cacioppoand Patrick indicate, health and wellbeing suffers as a result:

‘Given the importance of social connection to our species…it is all the moretroubling that, at any given time, roughly twenty per cent of people feelsuªciently isolated for it to be a major source of unhappiness in their lives…This nding becomes even more compelling when we consider that socialisolation has an impact on health comparable to the effect of high blood pressure, lack of exercise, obesity, or smoking.’ 127

A networks approach gives us a clearer understanding of thesepatterns of social inclusion and exclusion. In New Cross Gate, all theworkers consulted spoke of the importance of conducting outreach toexpand access to community activities among residents, especiallythose who do not easily come forward for services. People who do notspeak English as a rst language, ex-offenders, retired people, peoplewith low self-con dence, and people with literacy problems wereamong the groups cited.

124 Lowndes V, Pratchett L, and Stoker G, ‘CLEAR: Anauditing tool for citizen participation at the local level’,2006. http://www.ipeg.org.uk/papers/clear_sept06.pdf .The framework contends that participation workswhen people are involved in such a way that they: Canparticipate, L ke to, are Enabled to, are Asked to, and areResponded to when they do.

125 Halpern D, 2005, op cit.

126 Cacioppo J and Patrick W, l : H a naa h n s c a C c, 2008, p 15, WW

Norton and Co.

127 Ibid, p5

Applying network prinCiples to Community problems

http://www.ipeg.org.uk/papers/clear_sept06.pdf

http://www.ipeg.org.uk/papers/clear_sept06.pdf




In our research, one community worker argued that for practicalreasons, organisations have to focus on people on the margins, butnot the most excluded: ‘There are individuals and families so disgruntledand angry that they don’t want to be part of things, they put up thedrawbridge, so you have to look at families who are just on the edge.’

Other workers made a case for reaching the socially and economicallymarginalised through available family and friends and otherorganisations. A social network analysis has the potential to createsome leverage to reach those who are most excluded, because it canindicate ties through which the most excluded can be reached.

Some people are hardly connected at all. As gure six indicates, fromour survey in New Cross Gate we revealed that 5% of respondents hadno weak ties, while 2% of people were completely and utterly isolated.

Looking at all of these isolated respondents as a group, more thanhalf are older people (55+, with most of these 65+) while approximatelythree quarters have lived in the area for ten years or more, and,surprisingly, none are new arrivals to the area. Some sample quotationsfrom these respondents who are relatively isolated suggest that theirisolation is partly due to available amenities and chosen activities:

‘I used to use the local laundry until it closed – now I use the one inDeptford, so I tend not to see local people’ — r b ac A ca v ha a

n C ga .

‘The only activity I do is watch TV’ — wh b h a , v 75.

However, some isolation will be more to do with personal attributes:

‘I'm not a very sociable person, if I have to give a rating on sociability itwould be a zero’ — 35-44 a - a , v 10 a n C ga .

Those without any ties at all, 2% in our research, remain harder toreach, but at least social network analysis can highlight who andwhere they are, making it easier to direct resources to those most inneed. In the rst instance, once isolated people are identi ed byunderstanding their lack of networks, steps to help them make local

connections to individuals and groups can be identi ed (which maytake the form of befriending, and personal introductions to and bylocal people).

t c c n a a

1 p 6%

2 p ’ 7%

3 p ac , , ac v , a a 7%

4 2 a 3 5%

5 1,2 a 3 2%

f 6: i a a e c n C ga (n=280)



45

f 7: i a i v a as c s

i h a , h a ( a ) hav c .th v a a h

h ch a a / v ah a v a , c hc v ha a a

h c a h h h. th h a h a

, a a h h a h h ch h a ; ca

c c / c v c ( c c ), c/ac v /c ( c c ),

h ( c c ), a va v a( a c c ).

m h h h a h

c /c . o h a ,h 170 C p j c a a c a

-c c , a h s v

A v p a , h ndC, hba wa C a h A saC C .

sa ’ , h a h ndCa a, c a h

h , c h h a‘ ’. th h h h h sa ’ ac a a a

a c h ca – a a ac c

a g a v ’ a h

w a v ch h ha .

Community CAse study: untApped sourCes of supportin new Cross gAte





Community CAse study:kinsHip in tHe Contextof isolAtion

b c a a a c c a a c ac c a a .128 A c a

a a , ca h c a h

a h c aa .129 th ca a ha

ch v a ha v h , aha h a h ha v

a c aa .

i ca , h a h a h ac h a

a a c a h h ,, h a , a

c a va h a a – a ca ,c a va h c . th ‘ a v

c a ’, Ch a ’ , ha ac h h k h

p j c c h c ca aa v c , a h ac ah a h a

a h h a a a vc c c .

th h v h j c ,

c a ’ v cha a ha h a a v hh h c a h c

( h ). sch hh h h ch h a a ac a ; j c

a v c c h h h c ac a h c

h h a , c v a‘ a- ’.

i a h c a h h j c ca c a ha

a a ha c caac h c c .

s ch ac c ca h a-ca v h ha a c ac h

h a c c a h v a a a h v . th

a a h c a ac n C ga .

i c c h a ch, a hcha v h

h j c . V a a

h ‘ - v ’ c a a a h ac c . i c , a a ha

c h a v a ac h c a c a

a h j c ha , aa c a a , a a h h ca c . w a

h h a v a c a v h a .

f 8: r c a h hh

k :

Chpag a a



47

5.2 resilient Communities:reduCing frAgilityAnd mAintAining ConneCtiVity

r p a, in his seminal work on social capital, Bowling Alone,conjectured that a resurgence in civic participation might follow from‘a palpable national crisis, like war or depression or natural disaster.’The nancial crisis and subsequent recession may not be the scale of crisis Putnam had in mind, but the economic downturn invitesalternative approaches to building social capital and fosteringpro-social behaviour. 130 What the recession has wrought is the need fora deeper understanding of resilience, of our capacity to withstandsetbacks and unexpected shocks to the social and economic system.Resilience is an ecological concept and concerns adaptive capacity. It isnot about keeping everything the same, but about having some controlover the changes to which we are subject. Cs H de nes it as:

‘The capacity of an ecosystem to tolerate disturbance without collapsing intoa qualitatively different state that is controlled by a different set of processes.A resilient ecosystem can withstand shocks and rebuild itself whennecessary. Resilience in social systems has the added capacity of humans toanticipate and plan for the future.’ 131

By resilient we mean the capacity of a place or people creatively andsuccessfully to cope with and adapt to anticipated or unexpectedcircumstances that threaten to deplete existing nancial, social andcultural resources. We are beginning to make sense of how resiliencecan be understood in network terms.

The two major factors appear to be a) the strength or fragility of keycommunity hubs including their degree of interconnectedness, and

b) the extent to which networks forming to solve particular problemsremain constructively connected after a problem is solved. This latterpoint is related to the extent to which people value helping orbeing helped by neighbours more generally. The example overleaf from Knowle West provides a useful case study to illustrate some of these themes.

128 Christakis N and Fowler J, ‘th s a o a la s c a n ov 32 y a ’New EnglandJournal of Medicine, (26 July 2007) 357 (4): 370–379;Christakis, NA; Fowler, JH (22 May 2008). ‘th C c vd a c s a la s c a n ’(PDF).New England Journal of Medicine 358 (21): 2249–2258;Fowler, JH; Christakis, NA (3 January 2009). ‘d a cs a Ha a la s c a n :l a A a ov 20 y a h f a haH a s ’

.129 The case studies were selected from people supported

by the Kinship Project: an initiative funded by NewCross Gate NDC and the Child and Adolescent MentalHealth Services (CAMHS) team at the South London andMaudsley NHS Foundation Trust. The project works withchildren, young people and parents, directly supportingthose in need; building capacity in the local communityto recognise, and support mental health recovery; andbuilds networks to better integrate those working with theproject (back) into community life and support.

130 ‘m ac : i c b hav th h p cp c ’, Cabinet Oªce Discussion Document, Institute forGovernment, 2010.

131 Holling CS,‘r c a s a ec cas ’, Annual Review of Ecology and Systematics,1973, Vol 4.





i A 2009, a va c a , f b ,v h 89 , a a a a a

h h 90 v c . th 89 a 90a c v h a , a

c c , h h a -c c . wh h89 a a a a , h a a v c l Av , h ch

a a a hac a ªc h c a a a b .

th 36 v c a ava a , h q a / ªc

a , h ch a ah ,v c a h , h h ha ªc a h .

l ca l Ca , h ha c a c h va h

a :‘Something has to be done.’ l c ac ca c c , hha . sh a ach Ca Ha a ,

c k w m a C , aa h a .

Ca kwmC a a ac cha . t a v h ,

l h c ac , a h h h a h . sh a

c ac h ca a , h ev p .th ca C c a a

c f b a , aa v j a a .

A a c a h h, a a ,a h a a a va ;a a ch a cha , a c ’ a

h h a a ha . A c ’ b ;

a a c ’ 6a ah . o a c ’ v h

h a .

n h ha a a h .th c c a ha ‘it takes time ’ hCa Ha a ‘we need to capture thesestories’ , a c a a , C ad , c h . Ca ha

ha h a a a h , a ‘not just through the goodness of her heart.’

l c ac a h c aa h ca c c ’ a h .‘Being on your own - they shut their doors onyou,’ h a . sh a ach Ca a aa h a c .

A h h a a zca ach h ,

h a a c , a a kwmC,ac a c a . th hCa , da p a , mp b s h,ca h c a h a .th c a

c h ca a , a l c ‘ch c ’ h h c a

c ac A C c ( A uk).

l a C a v ca hh h v ha c h

h c a . da p a mp ha h f b a h v

a v c ha .C q , h c a ’ a a c

h l a da ’ ªc . l h h h v a a hh h 89 a a c .

f b a ca h l k wh h h h a h 90

, a h a , h h , c a a ha

h H a h C h ch a h , a ca c - a .

th c a ac h a h a a, a s 2009

h a a .

Community CAse study (knowle west): tHe strengtH AndfrAgility of A Community Hub in responding to CHAnge



49

This example serves to illustrate the value of having a community hub(Knowle West Media Centre) as a neutral and supportive venue, thepower of linking capital, in particular the connection between thedirector of Knowle West Media Centre and a local MP, the role of a shared narrative in bonding people together (a diverse range of vividstories on why the bus mattered), and the role of leadership (Lyn and Carolyn)and cooperation in a tangible expression of community resilience.The short videos used to lobby the bus company also provide a tangibleexample of socio-digital capital in action because they gave voice topeople who may not otherwise have had one.

In February 2010 First Bus withdrew the bus again because it wasonly being used by an average of seven passengers, and so far therehas been no major appetite to bring it back. Community networks aredynamic - they may shift, degrade, and recon gure as relationshipsform, break, mend and wither away. This example serves to illustratethat temporary social networks can be fragile, and need to strengthened,extended and diversi ed to enhance their social and personal valueover time. It remains unclear whether there is more potential energyin the social network that formed around the issue now than there wasa year ago. Did the processes that emerged from the removal of thebus create purely instrumental relationships, or was there an enduringshift in the social capital of the area?

We don’t yet know, but the most striking aspects of the case studywere a) Lyn’s remark that ‘Being on your own- they shut their doors onyou….’ and b) The role of the Knowle West Media Centre in coordinatingefforts to bring the bus back. It also did so while simultaneously addingvalue to its own enterprise by demonstrating the power of participatoryart and video to help local residents express their grievance. Giventhe importance of a hub in fostering resilience in this case, it is worthexploring the issue more generally.

However, given this case study and the outcome in terms of the busservice, we should remember the earlier point that social networksshould not necessarily be elevated on the basis of any structural,functional or moral claims to ‘rightness’ over other forms of socialphenomena. They may not necessarily function in enlightened andastute ways, either in terms of their process or in terms of theoutcome they ultimately seek.


Community networks are dynamic — they may shift, degrade, and recon gure as relationships form,break, mend and wither away.




i c , h -hac , a v (a

a ‘p ca ’ a a c ) a a ’ ca nC ga . th v a a a h ch c c

c ‘l ca p ’ h h -hac . p a h a v a

ac h h h ch c c c aac v ca c ca .

p h c ca h aac c a – h a c

– h c ha h h a a

ac c ca a aa va h c a a ha h c a . m v , v ca h a c a

h c a ippr a c ac ,

a c a h a h c aac a ha h h

c a a h j .132

s c , h ha a a c ac a

a c c h h n Cga . s h a v a h . th ac ha vc - ac v ha a cc : h ba wa C , h Asa C C , b H a hC a g sh /b sC ga .

o h h ca c , h – hba wa C (bwC) a bH a h C (bHC) – a ca

h s v , h A sa CC (AsCC) a g sh a ca

h Ha cha a k t ac v . o h , bwC, bHC a AsCC

a c -c c s v a k t a . g sh ,

a h , c -c c s v , k t a a m v .n a , h c v hw a h a a a a

h a a v ( ).

o h c , g sh (a ca a a ac v ) c c ac h

v a . A ac cac c hava a c a ac v

ca a a a a .

th ca c c a a h acc

‘c ’ h ba wa C . th ca c

h a v a a cc c (ndC, A sa h 170

C p j c ) a a c a

h ( v h a a h cac a h n C ga a a) aca h a h a a c a . th ba

wa C h h a hs v a a, a h c c a ca a

v n C ga , hba wa C a h

.

th a h ca ca c ch a

bwC, bHC a AsCC a a ca . wh h a v h c , ac c a a hav ca a ch

h ac a c h h. th ha v h

v ca c a a v c ha ca ch a n C ga ca

a a h h h a a h h h h

a h ava a . i a ha a

c - cac ca ( . .

a ) h c a v ah v v c a ac v a

ha a a h ca a aha a h - c a a c a v a a h ha h a a

c a a v c ac . i ,- c a a a a

a a c a a .

Community CAse study: Hubs, pubs And otHer sourCesof resilienCe in new Cross gAte



51

f 9: n C ga H





Community CAse study: neigHbourliness At tHe Coreof A resilient Community

A c h , a a c a ca , a va ha c c

h c c v . th a h a v c c a

c a va , c a c a aa a h h h v a

. f a a n C ga a a ha h h a h va

h a a h h c c . o a ch ha h va h a hav v c a . th

a ‘h h h ’133 a h h c

h ( ). th h h a h ha h a hav a -

c a a q n C ga .

i h a ca a c a a , h c c c ca hac , a c c a a, h

c v c a a h a- , a a

a h h ha h

h h h . i hc , h va c a

( ha a h ha ) aa a a . th a v

ha h cha ac a va v c ha a h .

H v , a ha a h a h h c a a

h ca ca a h , a a a ha h a

c a h v- h . f h

a , c a c ha c v a v

h c a c c .134

f 11 ca ha s v a t a h H h 31% a33% h c v , c h

’ ha h. s v a a hav a

h h , h ch ‘ ’ a h a a, a (c c a c ) c

h a a hav a a aªh ‘ h ’ hav c a h

h a a .



53

k (a a):

k t as vt a h Hw aHa chamm vH h

h

f 11: H b H n h , A a

f 10 H b H n h a C n





5.3 empowered Communities: ACCess,Coreness And Coordin Ation

We value shaping our circumstances as well as coping with andadapting to them. In this respect, the Department for Communitiesand Local Government has de ned empowerment as: ‘the giving of con dence, skills and power to communities to shape and in uence what public bodies do for them or with them.’ 135 Empowerment is increasinglybeing cast in network terms:

‘Sustained success in tackling neighbourhood deprivation will dependheavily on an approach which releases the capacity within thesecommunities themselves, which may include seeking to utilise the power of social networks.’ 136

The key elements of community empowerment appear to be:

a) ease of access to power and authority,) the nature of the network core in a particular geographic area and

c) the coordination of activity between organisations and citizens.

We will examine each of these in turn.

a) Access to power: inequality of power is network inequality

The ‘power’ in empowerment refers to the power to shape one’s ownlife in the context of others, rather than the power we have over others.This power has been described as ‘the production of intended effects’by Bertrand Russell and ‘the opportunity to lead lives they have reasonto value,’ by Amartya Sen. 137

Addressing inequality in power is the underlying motivation of muchcommunity development. In their research outlined in ‘The PowerGap’, Demos placed an emphasis on power as personal capability, butcommunity development has traditionally emphasised communityempowerment, in which power is viewed as an emergent propertyarising from access to and in uence with and through other people.In this sense power can be viewed as an emergent network property,and in principle comparing social networks could also affordcomparative measures of power. As philosopher Ha ah A stated:

‘Power is never the property of an individual; it belongs to a group and

remains in existence only so long as the group keeps together. When we sayof somebody that he is ‘in power’, we actually refer to his being empoweredby a certain group of people to act in their name.’ 138

Arendt’s claim is not the nal word on power, but it highlights thatsocial networks and the attendant social capital tend to be taken forgranted by those who have them. In this respect, the literature onempowerment makes an important distinction between subjectiveempowerment (the feeling of eªcacy) and actual empowerment(whether you are really changing things). We would argue that thosewho feel subjectively empowered on the basis of their own eªcacy donot realise that their empowerment is likely to be a function, in part,

of their social networks.

132 Muir R, ‘Pubs and Places. The Social Value of CommunityPubs’, 2009, IPPR

133 In terms of how this disposition was generated, it was 1of 15 multiple choice responses to a question about whattypes of groups and activities the respondent takes partin. Of these 15, ‘helping or being helped by neighbours’(which appeared towards the end of the answers offeredto the respondent and so should not be subject toquestionnaire design bias) was by far the most frequentlyreported – the next most ‘core’ node from this questionis ‘sports team or club’ which is ranked the 16th mostcore node in the whole New Cross Gate network. It isworth noting that those partaking in the pilot survey werenot given this multiple choice question and so this nodedoes not appear in any of the networks generated by thepilot data. As such, the ‘coreness’ of this node wouldundoubtedly be greater if the ‘core-periphery’ calculationwere applied to the second tranche of survey data only.

134 http://www.matthewtaylorsblog.com/socialbrain/eureka /

135 Communities and Local Government, ‘A Ac p a C e : b s cc ’,

October 2007.

136 Communities and Local Government, ‘s c a Ca a aC v c pa c a : A s a c r v Cp c ’, 2009

137 http://www.demos.co.uk/publications/the-power-gap ‘th p ga ’, Leighton D, 17-12-2009, Demos

138 Arendt H, o V c, 1970, Harcourt, Braceand Jovanovich.

139 Lerner M,s p : th p ch a c ev a l a h p ch i v a a

s c a t a a , 1998, Hb Publications.

http://www.matthewtaylorsblog.com/socialbrain/eureka

http://www.demos.co.uk/publications/the-power-gap

http://www.demos.co.uk/publications/the-power-gap

http://www.matthewtaylorsblog.com/socialbrain/eureka



55

i h , c n C ga v ca

a hav v acc :

A q a h v . th a h

43 ‘ a ’ ( a ach ) ha a aa a c h ha 12a v . o h 173 c h

a c a a a , 43 (25%) c aa h c a h h

h h a a) a h ) c h h c ac

h c , cha h ca .

b h c a , h a h a ca c

a h ha a c a c -h a . w a a

c c v c , h a; h c a a

h c a 12 (c c ). wh h c

ha a , h z ha h a h cha .

ev h c c , hav h c c a h .

o n C ga h h h h a a c c a

c a c v :‘Richer people,such as people from Telegraph Hill, know how touse the system. People from New Cross Gate tryto complain, but don’t know how things work,so they get frustrated and don’t get the resultsthey need.’

r h ch h h h c c , ch a

a h h b a h ndC h a a t a a r

A c a (trA), a h ca hH v , a ca a v , a nC ga a a a v

ch a c ac . th aa a c h h a

. i h a a v h h h ca

a a :

So someone comes to the TRA to get help to sortout something practical, like repairing thewindows, so I get on to the council and they saythey’ll sort it out in two weeks. The personcomes back to me after two weeks, then monthsand still nothing. Then he comes back and says

– you’re useless. The council lets people down,but I’m the baddie, they’re not.

th h ªc a . ma trA

a v h ha h a v h v . s ch a ca ha , a a

a ‘ ’. Av us ra m cha l :

‘Surplus powerlessness refers to the fact thathuman beings contribute to their existing powerlessness to the extent that their emotional,intellectual and spiritual makeup prevents them

from actualising possibilities that do exist.’ 139

l c a ha ch a c ca a ; q c - c c c a

a v c c c a c c ch - c a a .

Community CAse study: ConneCtion to powerin new Cr oss gAte

f 12: C c p n C ga





i h a a , h q a a . w c ha

a a a a v h a

. ra h ha a h h v ah , h -ac v

av h ha h c , h

a j c v , ah h , h a

j c v a j c v ,a c c . t av h

h a , a h c .

) th a ‘ a ’ h c

wh a c c ac nC ga , ha a ndC a v

ha h c a c c aa h h h a hav c c ’ a c .

th a h a a a ha a ca c

( 13) a a v a c z h a c c h

c c /v a c .140 thc h c c h c a

ac v a a ca , a cva a , h

ca a h , a

a a .

A a h h a h h a

a h ha , h va a ach a ac z . i c a , a ca c

a a a a h c c ,ch a h v c v a

c a ach , ac h h c - h

a – h v a c c ’ a c .th a h h c a

h a a h 14 a 15.

f 13: th C a p h n C ga



57

i c a ha h a a aa a n C ga h a ch

c c a h a aacc a a . th a a

ha a c c c ( h ), h ch acc v

h a c h hh ch a c aa cha .141 u , h h

ha ch h ndCa a - a c

ac v n C ga . b ha , h , h c ?

H , h c c h h c– h a ac ha a c

h ac h c – , a ca ha h a

h -c c c a h hh a . f 15 v a acc

h c h c

th h ha h h c c v a a h ha

a a , h a hac c a , h ch h

ha h a a a c a .

c) e h h c aa v a

C c a c a c c a a

h ch a ca c v hh c c v

Part of the empowerment message is thatcommunities are uni ed elds in which all issuesmingle. But it is also necessary to work from theadministrative structures of the different public service silos( a a c ha

j h ).

Community CAse study: wHo And wHAt is in tHe Corein new Cross gAte?


f 14: C n C ga( c a – h a c c )




Empowerment is best understood as anintervention from a pivotal point betweencommunities and public agencies to facilitateboth community strengthening and a moredynamic relationship between public servicesand their users. The job of empowermentcoordination is to nd leverage points both

within communities and agencies and enablethem to work together. 142

s c a a a ca v a ch v a h ch ca v

q c a a a. f a c , v v a h

a ach v . J a r c , h ca mp, a

q , h ch c aa c a v c ha h a ac v mp

h a a , ac v c h a a.t ca c c t a h H a a , h h n C ga , h c ‘ ca c c ’

a . A a h

c a v c ha c a a a acc a

c . o h a a c , h

v a , c ca q z a , a ca c , h ha

ªc a c , h a a h

h v a a v .

f 15: n C ga C d a

k :

i v a (private)th c c a ap c c a a o h (organisation, group, disposition)



59

Summary

This section has examined three concepts that appear particularlyrelevant to communities and policy initiatives in the currentclimate: support, resilience and empowerment, and made some earlysteps towards making them work in network terms, as well ashighlighting the opportunities for the design of better communitydevelopment strategies.

• The importance of mapping bridging nodes and organisations wasidenti ed as critical for building network structures that offermutual support and reduce isolation and loneliness.

• The strength and variety of hubs and the propensity for networkdecay were considered key aspects of resilience.

• Network position, the nature of the network core, and the degree towhich local organisations are coordinated appear to be key aspectsof empowerment.

Now that we have an account of what a connected community mightlook like, and the areas (inclusion, resilience and empowerment) wherecommunity policy could bene t from applying a network-based approach,we can consider how to they can be shaped and built in practice.

140 The black centre of the core/periphery ‘eye’ indicates thedensity of connections at the core.

141 See for example Krebs and Holley, ‘b s aC h h n w av ’,http://cfc.barrexternal.org/wp-content/uploads/2008/10/building-smart-communities-through-network-weaving-

krebs-holley-2006.pdf

142 ‘Empowerment Skills for All’, PACES.’


http://cfc.barrexternal.org/wp-content/uploads/2008/10/building-smart-communities-through-network-weaving-krebs-holley-2006.pdf









The question is, can we changethe value of a street corner in theway we can change the value of shoe polish? And the answer is yes.

— k v s av147

143 Crow G,‘s c a n a s c a e c , a

v v h a , ph a ( ) s c an a s c a e c : s c ca a p cp c v, 2004, Ashgate.

144 For a more detailed account of meta-networking, seeGilchrist A, C c a Ca a , 1998, SCCDNews(18)

145 Krebs and Holley 2006,http://www.orgnet.com/BuildingNetworks.pdf

146 Ibid

147 http://www.thersa.org/events/vision/vision-videos/kevin-slavin-playing-the-city

seCtion 6: building ConneCtedCommunities

If we accept the argument that social networks may be a usefuldiagnostic and explanatory tool, what can we do with this knowledge?

As Crow indicates, networks are just a starting point:

‘The diagrammatic representation of a social relationship in terms of linebetween two nodes in a network is necessarily only the beginning of thesocial scienti c inquiry into the signi cance of this connection. Researchershave to nd out precisely what passes along these lines before pronouncements about the nature of the network can be made.’ 143

In section three we argued that the process of measuring networks andre ecting them back to communities as sociograms was a potentiallyimportant and underexplored aspect of community regeneration. Wealso believe that the research process itself should be a key component

of any regeneration strategy. While network structures and functionsmay be dynamic in certain ways, it is the process of measuring andmapping social networks that has the potential to shape thenetwork in constructive ways, especially when the research is doneby local participants.

The key components of any strategy that seeks to build social capitalare the effective use of existing social networks, and the shaping of new ones. At a minimum, use of existing networks requiresunderstanding the connectivity of key nodes in that network, whilethe shaping of new networks requires skills of network weaving(deliberately building relationships and supporting collaborations

between people, and between people and organisations) and meta-networking (ensuring that the social, psychological and geographicconditions are suitable for people to build, shape and sustainnetworks). 144

It is important to appreciate the pivotal role of the researcher in raisingparticipants’ awareness of their connectivity, and thereby directlyin uencing the way existing connections are used and futureconnections are formed. The research process itself can be a formof networking and network weaving.

6.1 network weAVing

Community consultants Krebs and Holley 145 show how social networkscan be used to build social capital. They suggest that weavinga network requires two iterative and continuous steps of knowing thenetwork, for example taking regular snapshots of it to evaluate progress,and weaving the network, which requires following the four phasenetwork knitting process outlined in gure 16 .


http://www.thersa.org/events/vision/vision-videos/kevin-slavin-playing-the-city







building ConneCted Communities

61

Krebs and Holley’s four-stage process is outlined and illustrated below:

1 sca f a require network weaving to move to the nextstage, Hub-Spoke.

2 H -s means moving from single to multiple network weavers(and hubs and spokes) and moving from network weaving tonetwork facilitation.

3 m -H Small World Network develops weak ties across clustersand strengthens the network; political and ‘turf’ issues can act asbarriers at this stage.

4 C -p h is the ideal state, requiring network maintenanceand bridges to other networks, while the periphery reaches intonew areas and weavers focus on core projects for maximum impact.

The end goal for vibrant, sustainable community networks, thecore-periphery model, emerges after many years of network weavingby multiple hubs. While such a growth in connectivity may be naturalin healthy communities, network weaving is the key to using socialnetworks to develop social capital. According to Krebs and Holley:

There are two parts to network weaving. One is relationship building,particularly across traditional divides, so that people have access toinnovation and important information. The second is learning howto facilitate collaborations for mutual bene t. Collaborations can varyfrom simple and short term – entrepreneurs purchasing suppliestogether – to complex and long-term – such as a major policy initiativeor creation of a venture fund. This culture of collaboration createsa state of emergence, where the outcome — a healthy community —is more than the sum of the many collaborations. The local interactionscreate a global outcome that no one could accomplish alone. 146

Throughout this network weaving process, network maps shouldguide the way because they reveal what we know about the networkand uncover possible next steps for the weaver. What Krebs and Holleydo not make explicit is just how diªcult measuring social networkscan be, and although their linear model is useful, network connectionscan sever as easily as they form, often due to contentious issues thatlead to network divisions.

6.2 Contested spACes And ‘turf issues’

The design of social spaces has a direct bearing on the nature of socialnetworks that form, and in New Cross Gate we noticed at least three areasthat raised interesting issues in terms of the growth and decay of networks:

1) Putting the adventure back in to the playground

One respondent remarked that local divisions between groups of youths in the area are de ned in relation to the Somerville AdventurePlayground; there is the ‘Somerville side’ of the community and the‘Monson side’ of the community. This division is arguably evident inthe network we uncovered, because no residents from the Monsonarea north of the Kender Triangle cite the adventure playground aspart of their network.

f 16: k a H n w av

1

2

3

4




Somerville Adventure Playground serves to illustrate the point madepreviously about social goods (see section 2.1) because the playgroundis not currently functioning as a public good due to network effects.A community resource that should serve to bring different parts of thecommunity together has instead become a boundary line, keepingpeople apart. Younger people have strong feelings that the playgroundshould be a priority for investment, and more generally respondentsargue for investing in informal spaces where residents from differentparts of New Cross Gate can mix, with provision for younger peoplefelt to be a particular need.

2) ‘Trying something new’ at Sainsbury’s

Sainsbury’s is clearly linked to many respondents in the network. Thisfact is signi cant, because it raises the potential that Sainsbury’s, theonly supermarket in New Cross Gate, could act as an important bridgebetween employed and unemployed clusters in the local network.Linking these groups is important to the development of areas such asNew Cross Gate where unemployment levels are high, asGranovetter’s work suggests. This might be done not through, say,a formal job fair in the Sainsbury’s café, but by building space andopportunity for people from different socio-economic backgrounds tointeract regularly and form connections and knowledge of each other.We believe it is precisely efforts to make such links which could havea transformative effect on an area, and that it is an opportunity for theproprietors of such hubs to make them available for this purpose,at least some of the time. In such ways, the private sector becomes

a co-participant in the Big Society.

Our network ndings also suggest that there should be more

community involvement, engagement and say in ongoing Sainsbury’splans to regenerate the local area, 148 and that locating communityservices within this development could be an effective move, given thenetwork centrality and potential role of Sainsbury’s.

3) Post Office possibilities

While conducting our research, further possibilities for encouraginga diverse range of people to meet on a regular basis were discovered.For example, the considerable space outside the New Cross Gate PostOªce might be better used to encourage network building, particularlygiven the site’s proximity to transport hubs and the consequently high

footfall past the area. Places such as a bench, a fruit market, or a giantpublic chess board would lead people to linger and talk. However,while the intelligent use of public space has an important role to play,network weaving requires skilled people more than anything else.

...while the intelligent use of public space has an important roleto play, network weaving requires skilled people more than

anything else.



63

6.3 Community orgAnisersAre relAtionsHip builders

The coalition government is committed to increasing the number of active community organisers. It is important that such organisers takea leading role in network weaving and are encouraged and trained todo so. However, as a recent PACES 149 report indicates, the form of funding for such people is absolutely pivotal. Indeed, in New CrossGate we found that another perceived barrier to effective communitynetworking was the way that funding is structured, what is classed asan outcome, and how these outcomes are measured, with theresult that, as one respondent commented, ‘ connecting people is notmeasured in the funding we get, maybe there needs to be more scope for things like this.’

As another worker put it, ‘ The funders are output-driven, organisationsare so busy delivering these outputs that they don’t have the capacity for

partnership.’ The way that the local organisations perceive and relate tocoordinating and funding bodies such as the NDC is important forunderstanding the context of community development and drawing outlessons for how to move forward. Moreover, one community workerencapsulates a widely expressed viewpoint:

‘Oh, we all jolly along ne, but it could be much better. At the end of theday, people feel like they could be competing for funding. That’s the elephantin the room.’

Several community workers spoke about the strong relationshipsbetween a group of well-established community organisations, incomparison with more forced relationships with others organisations

they de ned as having ‘parachuted in’ in order to obtain NDC funding.

6.4 networking is not A four-letter word

’Networking’ appears to have been taken from North American soiland transplanted, uncomfortably, into British parlance, without anysupporting context. As a term, it can sound calculating andinstrumental to British ears. We need to get beyond this limitation.The key to addressing turf and funding issues is to recognise the valueof networking more explicitly, and the crucial importance of meta-

networkers and network weavers. As A g chindicates:‘Perhaps the most important, although somewhat tautological, function of networks is their capacity to support networking: enabling people to shareideas, consolidate relationships, exchange goods and services, and co-operate.’ 150

However, as the Krebs and Holley model suggests, networking requiressupport and reinforcement, which means recognising the communicationchannels within communities, especially those that span boundariesand bridge schisms. 151Such abilities are often embodied in individuals,variously termed Switches 152, Social hubs 153, Linkers154, Weavers 155, Criticalnodes 156, Community catalysts 157, or Meta-networkers 158. Regardless of the

label, what these key people do is put other people’s networks together.159

148 http://www.sainsburys-newcrossgate.co.uk/index.html

149 Chanan G and Miller C, 2010, op cit.

150 Gilchrist A, 2009, op cit.

151 Ibid, Gilchrist also provided references for Switches,Social hubs, Linkers, Weavers, Critical nodes, Communitycatalysts, and Meta-networkers.

152 Castells, The Internet Galaxy, New York, 1996, OxfordUniversity Press.

153 Gladwell M, The Tipping Point: How Little Things Canmake a Big Difference, 2000, Little Brown.

154 Fraser et al, ta e s c a Ca a –r ch n , p C c , Department forInternational Development, 2003.

155 Traynor B, Community Building: Limitations and Promisein De llipis J and Saegert S (eds), th Cd v r a , 2008, Routledge.

156 Dale A and Sparkes J, Protecting Ecosystems: NetworkStructure and Social Capital Mobilization, Cd v J a, 2008, Vol 43.

157 Creasy et al,ev n g n h ? A h l b p c pa c a a C

C h , Involve, 2008.

158 Gilchrist, A, 2009, op cit.

159 The importance of such people is currently contested.See for instance i h t p t a ?by DuncanWatts: http://www.fastcompany.com/magazine/122/isthe- tipping-point-toast.html


http://www.sainsburys-newcrossgate.co.uk/index.html

http://www.fastcompany.com/magazine/122/isthe-%20tipping-point-toast.html




http://www.sainsburys-newcrossgate.co.uk/index.html




We believe this function can be performed bypeople who are already widely known but donot have community networking as part of their recognised skill set, including postmen,dustmen, lollipop ladies and pub quiz masters.If they are to become effective communityorganisers, they will need to develop someof the skills indicated by Gilchrist, includingaffability, integrity, audacity, adaptability,tenacity and the ability ‘to make rapid andsophisticated appraisals of complex anddynamic processes from their observations of informal interactions.’ 160 Gilchrist adds that:

‘Networking is effective for communitydevelopment because it is personal, involving morethan super cial connections devoid of emotionalcontent. Networking is not about exploiting contacts in a manipulative or sel sh way, butabout establishing levels of trust, goodwill andmutual respect that run deeper than a sporadic and perfunctory exchange of information,business cards or favours. Personal relationshipsmake it easier to make requests and suggestions,especially when these are inconvenient,complicated or hazardous.’ 161

It seems obvious that some will be better at such networking thanothers, and as Gilchrist argues, networking can be considered a formof expertise:

‘Experienced community development workers develop the art… over timeand nd it invaluable. Intimate knowledge based on networking covers suchareas as who gets on with whom, who used to work for which organisationand why the director of one local organisation has the ears of the chair of social services. Gossip is among the most precious information incommunity work. Such material is too sensitive and too complex to store ona computer…what a competent community development worker carries inher head is a highly sophisticated relational database.’ 162

However, networking has its limitations too. We have to be particularlycareful to guard against what journalist s z d a has called‘coercive participation’ and ensure at all times that we do not use the

ndings from social network analysis as injunctions to force peopleto act against their will. Indeed, as A w aargues:

‘The notion that communities can be orchestrated into existence displaysa casual disregard for the very thing that makes them special: communality– a voluntary association of interested parties.’ 163

f 17: n h g( a )



65

6.5 women’s work?

Another signi cant dimension to emerge from our data is gender,given that it is commonly held that in communities women do a lotof the social capital building ‘work’. 164 Women are hugely active inweaving together and mediating between their neighbours andcommunities. Figure 17 shows that women value neighbourlinessmuch more than men.

This nding corroborates recent research on online networks whichshowed that more women than men took part in online socialnetworks, and were more active users. 165 On a related point, in ourresearch we found that both women and men generally have similaraccess to resources (although women appear to have few connectionsto the media and to people who can hire people from time to time),but access them in different ways. Women are more likely to accessresources through their local social connections, whereas men reportbeing more self-reliant, believing themselves to have, or know howto acquire, the capabilities they need to address their own problems.Gilchrist explores another nding regarding gender:

‘In many informal networks there often seems to be one individual who keepsin touch with the others, who arranges get-togethers, has up-to-date newsand contact details, and generally ensures that everyone stays on more or less good terms. In families this role is often played by women, and there isevidence that women’s emotional labour creates and maintains networkswithin other social settings, such as the workplace or within communities.’ 166

summAry

In section six we have argued that building connected communitiesseems to bene t from at least three inputs:

• Participative social research

• Networking

• Network weaving

None of these elements have featured in recent discourse about theBig Society, which is why it is important to highlight them now.

I think the big society (in lower case) will inevitably outlast theconcept which is the Big Society.But the latter could be a lifeline

for activity that’s really importantat a dif cult time.– J a d 167

160 Gilchrist A, 2009, op cit, p92

161 Ibid, p141

162 Ibid, p62

163 Williams A, ‘Who Needs Community Anyway?’ Inth f C : r a d a h g ae a a . 2008, p7, Pluto Press.

164 See Bruegel I, Social Capital and Feminist Critique, inFranklin J (ed) w a s c a Ca a , w pa

12, London: South Bank University, 2005; dl, w a C Ac (2nd edition) 2006,Bristol: The Policy Press.

165 Alemain A, Wartman K,‘o s c a n Ca : u a wha ma s

C ’. Dec 2008, Routledge.

166 Gilchrist, A, 2009, op cit, p97.

167 http://livingwithrats.blogspot.com/2010/05/building-big-society-in-big-society-way.html


http://livingwithrats.blogspot.com/2010/05/building-big-society-in-big-society-way.html







seCtion 7: ConneCted Communitiesin tHe big soCiety

The coalition government’s vision of the Big Society is not entirelyclear, but according to Prime Minister David Cameron it is based on

the premise that ‘we can all do more’. In this sense, the Big Society isa political injunction to build a big society (lower case) wherein oursocial and civic capacity grows to solve problems that were previouslyaddressed by the state. Social capital is the currency of this big society,and social networks hold the reserves of that currency. At the sametime, this should not be a reason to withdraw support and funding fromdeprived areas, particularly those among the most deprived neighbourhoodswhose entrenched disadvantage appears particularly acute.

Government advisor Lord Wei analyses the 'ecosystem’ of the BigSociety in a manner that calls for a deeper understanding of socialnetworks at a community level. 168 For example, one of the keypriorities of the Big Society policy agenda is to recon gure policydesign and delivery, which is precisely the promise of a social networkapproach. Moreover, one of three key objectives of the Big Society is toget people ’more involved‘ in their communities, in order to ’reduceisolation‘ and ‘create stronger social ties’. These objectives can all bemeasured in network terms, which is signi cant because Lord Weialso indicates that there will be greater local discretion to use differentapproaches of measurement and commissioning. 169

The growth of social capital will be one of the metrics used to indicatewhether society is getting ‘bigger’ and deciding on the level of geographical analysis to measure progress is also indicated as important.In this respect we believe that urban areas of high population densityof roughly a square mile, like New Cross Gate, will lend themselveswell to social network analysis because they are big enough to comprisea diverse range of social assets but small enough for the connections of an unbounded sample to be meaningful to participants.

The recent Cabinet Oªce document, ‘Building the Big Society’, includesa noteworthy conceptual point that re ects this understanding. Thedocument made reference to ‘ Society – the families, networks, neighbourhoodsand communities that form so much of the fabric of our everyday lives.’ 170 Gabriel Chanan and Colin Miller’s recent report: ‘The Big Society:How Could it work?’ includes the insightful subtitle : A positive idea atrisk from caricature. Families and neighbours are complex, but they areperceived in broadly positive terms. ‘Communities and networks’, on theother hand, need to be made more tangible and vivid if ‘the Big Society’is going to be made real.

RSA Connected Communities offers a perspective that illustrates whya Big Society cannot be monolithic on a national level, and will beeasier to achieve with some people, in some areas, simply becauseour ability to engage socially and participate civically is directly relatedto the resources in our social network. Indeed, we believe that if weare not careful about how to build the Big Society, network effects willperpetuate some of the inequalities that civic and political participationare supposed to redress.

168 Lord Wei: ‘Building the Big Society’, Institute forGovernment, July 6 2010

169 Ibid

170 See http://www.cabinetoªce.gov.uk/newsroom/news_releases/2010/100518-news-big-society-launch.aspx

171 David Cameron: The Big Society: Hugo Young Lecture:Nov 10 2009.

http://www.cabinetoffice.gov.uk/newsroom/news_releases/2010/100518-news-big-society-launch.aspx






ConneCted Communities in tHe big soCiety

67

The Big Society, as envisaged by the coalition government, thereforedepends upon good social networks, where ‘good’ is understood tomean links to important sources of information and power, and accessto people and institutions that offer skills and resources that arerelevant and meaningful to the groups and individuals who seek themout. As we have indicated above, such connections can, at least inprinciple, be created or woven, and their growth can also be measured,providing a Big Society metric that tells us things about social capitalthat existing secondary data analysis cannot.

However, to get a feeling for what ’Big’ is, we also need to be clearabout what ’society‘ means in terms of network scale. It is easy to slipfrom talking about ‘neighbourhoods’, to reminiscing about ‘communities’,to lamenting the various ills of ’society‘. Such terms will always befuzzily de ned and contestable, and at each level of scale we needdifferent kinds of measurement.

At the very least, social network analysis allows us to make the ideaof the Big Society more tangible and visual, but we also thinkthat a network approach can inform existing ideas on making theBig Society a programme for government.

7.1 wHy does group membersHip mAtter?

One of the easiest metrics by which the growth of the Big Societymight be measured is group membership, and social network theoryvalidates the importance of group membership. This lies not merelyin the activities of the group, but in the fact that most group memberswill be members of other groups, so joining a group potentiallyprovides access to numerous network bridges that would otherwisenot have been available (see section 4.1, Law 5 on hyperdyadic spread) .

7.2 diVerse network sHApes in tHe uk’s93,000 sQuAre miles

The existing Big Society Network, led by Lord Wei, has a ‘square mileinitiative’ and seeks to encourage people to make a social contributionto whichever square mile they live in (of which there are 93,000).While there is much to commend in this idea, it is important not torepeat the mistakes that stem from an exclusively geographical

emphasis of belonging, as outlined in section 2.2 earlier. However,a useful starting contribution would be to attempt to measure theexisting social networks in any given square mile, as we attempted inNew Cross Gate.

David Cameron has said that he wants to ‘ use the state to remake society’ .171 Our research suggests that the state may have an unconventional roleto play in providing support for community researchers, whether inthe form of providing a standardised research instrument to makedata collection easier, or centrally analysing the data that is collectedlocally. Such research would not only serve to create practicableknowledge, but is also in itself a form of political participation and

collaboration in keeping with Big Society principles.

Social capital is the currencyof this big society, and socialnetworks hold the reserves of that currency.




7.3 tHe relAtiVe effiCienCy of stAte And soCiety

One of the premises of the Big Society programme appears to be that,compared to the state, civil society makes relatively eªcient use of available resources. In so far as the Big Society is supposed to increaseeªciency at a local level, such eªciency depends upon coordinatingindividuals, groups, third sector organisations and local government,and such coordination is de nitely facilitated by understanding howthese different stakeholders interact in network terms.

While we do not expect the rst wave of the Big Society to involve socialnetwork analysis on a national scale, we do think our work makes a casefor attempting to reframe the idea of eªcient resource allocation asa metric that is informed by social network analysis. For instance, if the government is serious about its intent for every citizen to havesome form of group membership, the most eªcient way to achieve thisis not yers through doors of unequal importance, or local noticeboardswhich provide information but do not lend themselves to imitation.

A more promising and eªcient approach is to seek to create socialcontagion, which can usefully be informed by the relative centrality of different parts of the network. For instance, those who are particularlyisolated and disengaged are best accessed through whichever peoplethey are connected to, however loosely, rather than by the injunctionsof strangers, however well meaning.

7.4 Community orgAnisers

A central pillar of the Big Society programme for government is the

intention to create a ’Neighbourhood Army‘ of 5,000 full timeprofessional community organisers. 172 This idea is very powerful, butthe quality of such an army, how it is recruited and trained, deployedand valued, matters as much as the quantity, and like any army it mayrequire some level of strati cation to work effectively. In this respect,Chanan and Miller argue:

‘What is needed to make the big society work is not simply more communityworkers or organisers, but a wave of ‘super’ community developmentworkers with the status and training to be able to mobilise and coordinatea myriad of contributions to citizen empowerment from other voluntaryand statutory front-line workers. This means that these individuals will

have to be credible community involvement leaders in the eyes of other localworkers. They will have to be genuinely experienced and gifted in this complex and demanding eld…This is where the transformative potential lies...’ 173

Our research gives some pointers to this transformative potentialbecause the ‘super connectors’ that Chanan and Miller highlight canbe identi ed through existing patterns of connectivity. More generally,we suggest that such community organisers should be given basicinstruction in social network analysis as part of their training, and thatthe government could have a role in coordinating the analysis of thelocal data that is collected.

While we do not expect the rstwave of the Big Society to involvesocial network analysis ona national scale, we do think our

work makes a case for attempting to reframe the ideaof eªcient resource allocation asa metric that is informed by socialnetwork analysis.



69

Social network research should not be viewed as a prelude to plannedcommunity action, but more as an integral part of the ongoing processof fostering pro-social behaviour. Our contention is that socialnetwork research, in and of itself, encourages those who take part toconceive of themselves in network terms, and thereby re ect on theirresponsibility to use, sustain and develop their networks.

Afterword:

wHAt next for ConneCte d Communities?

Over the next few months the RSA will be working with communityorganisations and residents to co-design ideas for social networkapproaches to community regeneration that require relatively low-costresources. We will attempt to take a strategic approach. What doesa social networks strategy for a community look like, and how does itinform and complement public service delivery? Ideas will beprioritised with local people and we will seek to test them in practiceover the coming year. We believe the visualisation of individual andcollective social networks is potentially important and we hope to testthis assertion in our practical work. In Knowle West we will becontinuing to collect, stimulate and analyse data on socio-digitalinclusion, and plan to publish a report on the subject later this year.

Following these next phases of work in New Cross Gate and KnowleWest, we will seek to integrate the learning from the RSA’s SocialBrain project 174 with our Connected Communities programme toexplore how methods to improve individual behaviour, decision-making and re ection can be networked to amplify their impacts.

The RSA will soon be launching two further programmes under ourConnected Communities banner. Firstly, we will be commencinga longitudinal study to test the effectiveness of social networkapproaches to mental wellbeing and social inclusion. The work willbe delivered in seven action research sites in England in partnershipwith the London School of Economics and the University of CentralLancashire, and a range of local stakeholders. Both New Cross Gateand Knowle West are involved in the programme, and the latterwill explore, in particular, the role of online networks. The work issupported by the Big Lottery Research Programme.

Secondly, we will be launching a Social Capital Innovation Networkthrough which those interested in developing social capitaland network approaches to social and economic problems cancollaborate, share ideas and experiences, access research tools andanalysis, and draw on the wider work of the RSA’s ConnectedCommunities programme.

f a , . h a. / j c /c c -c c ac s v b , d c r a ch av . @ a. ..172 http://www.conservatives.com/News/News_stories/2010/

03/Plans_announced_to_help_build_a_Big_Society.aspx

173 Chanan G and Miller C, Spring 2010, op cit.

174 http://www.thersa.org/projects/social-brain

ConneCted Communities in tHe big soCiety

http://www.thersa.org/projects/connected-communities

mailto:[email protected]

http://www.conservatives.com/News/News_stories/2010/03/Plans_announced_to_help_build_a_Big_Society.aspx






mailto:[email protected]

http://www.thersa.org/projects/connected-communities



71

Appendix: reseArCH QuestionnAire

Hello, I am doing some work for the Royal Society of Arts Community ResearchProgramme. This questionnaire is for a project that looks at how people connectwith others in and outside New Cross Gate as part of their daily lives. The ndings of the project will be used as the basis for a strategy designed to improve communityrelations in the area. In the questions I will ask you full names of people andorganisations that you know so that we can produce maps of these connections[SHOW RESPONDENT EXAMPLE MAP].

The information you provide about yourself will be kept con dential and will notbe shared with anyone outside the project. We will ask you for names of peopleand organisations that you know, but no information you give will be used ina way that can identify you or anybody else you mention in this survey. When wecome to produce our ndings all of the names that you give us will be coded soas to keep this information con dential.

i v a

i v a s a(if applicable)

i v a H / a .(if applicable)

r a

s c A:

p a , va c va , a a na 1 na 2 na 3 na 4 na 5 na 6

.e a q wh h a cha ? (p a a a a )

J h s h Ja s h J h J Ja J J b J w a

.e a q o a ca 1 5, h va a a

ach h , h 5 v va a a 1 a c a va a ?

3 4 2 5 5 1

.e a q wh ch h v

ca a a?

Q1A a a , h a h

j c a h? (p a a a a )

o a ca 1 5, h va a a ach h

, h 5 v va a a 1 a c a va a ?

wh ch h v ca a a?

A a a c

Q2.wh a h ha a v c

c a a ? (p a a a a )

o a ca 1 5, h va a a ach h , h 5 v va a a 1 a c a

va a ?

wh ch h v ca a a?

A a a c

Q3.

i a ha h a a a , a . i ch a a h ca

( a a ) av ? (p a a a a )

o a ca 1 5, h va a a ach h , h 5 v va a a 1

a c a va a ?

wh ch h v ca a a?

A a a c



Q4. wh c h v a a h ? (p a a a a )


a c a va a ?

wh ch h v ca a a?

A a a c

s c b: ’

na 1 na 2 na 3 na 4 na 5 na 6

Q5.wh a h h a h h ? (p a a a a )


a c a va a ?

wh ch h v ca a a?

A a a c

Q6.

C h ha h aq , h c h c ha

h ? (p a aa a )


a c a va a ?

wh ch h v ca a a?

A a a c

Q7.wh ha c h c ac a h

h c , chah ca ? (p a a a a )


a c a va a ?

wh ch h v ca a a?

A a a c

Q8.

[giVe sHowCArd 1 to respondent] f h , a h ca h h

q a a a h , aa a a h? p a h v

h h ca c ach .(C c a a a )

1 Va : 2 Va : 3Va : 4 Va : 5 Va : 6 Va :

7 Va : 8 Va : 9 Va : 10 Va : 11 Va : 12 Va :


a c a va a ? (p a a a h va a a )

13 Va : 14 Va : 15 Va :

A a a c

72



73

s c C: ac , , ac v a a a

na 1 na 2 na 3 na 4 na 5 na 6

Q9.

[giVe sHowCArd 2 to respondent] u h ,a ha a ac v a

a . p a h h h cac ach /ac v . (C c a a a )

1 Va :l ca

2 Va :l ca

3 Va :l ca

4 Va :l ca

5 Va :l ca

6 Va :l ca

o a ca 1 5, h va a a ach h ac v , h 5 v va a a 1

a c a va a ?

7 Va :l ca

8 Va :l ca

9 Va :l ca

10 Va :l ca

11 Va :l ca

12 Va :l ca

wh ch h ac v a ca ?(d a a a )

13 Va :l ca

14 Va :l ca

15 Va :l ca

A a a c

Q10.

i a h ha c c h h h , h

? (p a a a ac -speCifiC nAmes of plACes reQuired)

A a ca 1 5, h va a ha ach hac c , h 5 v va a a

1 a c a va a ?wh ch h ac a ca ca a a?

A a a c

Q11.wh h a

ha ’ ca a a? (p a aa ac a )

A a ca 1 5, h va a ach hac , h 5 v va a a

1 a c a va a ?

wh ch h ac v , aca , ca a a?

A a a c

Q12.

[giVe sHowCArd 3 to respondent] d v , hav c v , a h

ca a a c ca a a h ca 3? p a h v hh ca c ach a a c .

(C c a a a )

1 Va : 2 Va : 3 Va : 4 Va : 5 Va : 6 Va :

7 Va : 8 Va : 9 Va : 10 Va : 11 Va : 12 Va :

o a ca 1 5, h va a a ach ha a c , h 5 v va a

a 1 a c a va a ? (p a aa h va a a )

13 Va : 14 Va : 15 Va : 16 Va : 17 Va : 18 Va :

19 Va : 20 Va : 21 Va :

A a a c

s c d: c a ava a

I am going to hand you a list of skills and resources that someone you know, or you yourself, may have.I would like to know if any members of your family or friends have these skills or resources? And are thereany less well-known acquaintances of yours that have these skills or resources? By acquaintance, I meansomebody that you would have a conversation with on the street if you bumped into them and whose

rst name you know. I would also like to know if you have these skills or resources yourself.

Please hand Showcard 4 to the respondent. Begin by asking (i.) whether the respondent has a familymember with the respective skill or resource. If ‘YES,’ move on to ask (ii.) if the respondent themselveshas the skill or resource. If ‘NO,’ ask them about friends. Only if the answer is ‘NO’ again, ask if theyhave an acquaintance with the resources or skills. Please note, if the respondent tells you that somebodyis both a family member and a friend then the person mentioned should be counted as a family member.



s c d: c

d a a / / acq a a c h … n fa f Acq a a c

…a a h … … 1

…ca a a , ca h h a a c ? [0] [1] [2] [3] [4]

2 …ca a a a a a ? [0] [1] [2] [3] [4] 3 …ca h ? [0] [1] [2] [3] [4] 4 …ha v ? [0] [1] [2] [3] [4] 5 … a h ca c c ? [0] [1] [2] [3] [4] 6 … a a h ? [0] [1] [2] [3] [4]

7 …ha c ac h a a , h a tV a a ? [0] [1] [2] [3] [4] 8 … a a ? [0] [1] [2] [3] [4] 9 …ha a a c a a ( a , )? [0] [1] [2] [3] [4]

s c e: a a

e1 p c e2 A (circle number of years) 16-18 18-24 25-34 35-44 55-64 65-74 75+ e3 g (circle one option) ma f a e4 e h c (circle one option) wh b h b ac b ac b h Ca a m

wh i h b ac b ac b h A ca o h

wh o h A a A a b h i a d c

Ch A a A a b h pa a

e5 fa /h h a (tick one option) ma v h a ,no ch ma v h a ,with ch

s a a

s h h

o h - h h

e6 l h v n C ga(circle number of years) +1 2-3 4-5 6-10 10+ e7 e a (tick one box) f - a V a

pa - a u

f - ca / a o hpa - ca / a

e8 H a (tick one box) C c a sh h

p va n a

H - o h

H

s c f: ca c ac ?

f1 Ca ch h a - - v ? y n

f2

w h a a h h ?y ni h / h, h ca ach ?

ph e a

74

connected communities - how social networks power and sustain the big society

Documents