understanding social capital in recently restructured urban neighbourhoods: two case studies in...

http://usj.sagepub.com/Urban Studies

http://usj.sagepub.com/content/44/5-6/1069The online version of this article can be found at:

DOI: 10.1080/00420980701256047

2007 44: 1069Urban StudReinout Kleinhans, Hugo Priemus and Godfried Engbersen

Case Studies in RotterdamUnderstanding Social Capital in Recently Restructured Urban Neighbourhoods: Two

Published by:

http://www.sagepublications.com

On behalf of:

Urban Studies Journal Foundation

can be found at:Urban StudiesAdditional services and information for

http://usj.sagepub.com/cgi/alertsEmail Alerts:

http://usj.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://usj.sagepub.com/content/44/5-6/1069.refs.htmlCitations:

What is This?

- May 1, 2007Version of Record >>

at MEMORIAL UNIV OF NEWFOUNDLAND on August 3, 2014usj.sagepub.comDownloaded from at MEMORIAL UNIV OF NEWFOUNDLAND on August 3, 2014usj.sagepub.comDownloaded from

http://usj.sagepub.com/

http://usj.sagepub.com/content/44/5-6/1069

http://www.sagepublications.com

http://www.urbanstudiesfoundation.org/

http://usj.sagepub.com/cgi/alerts

http://usj.sagepub.com/subscriptions

http://www.sagepub.com/journalsReprints.nav

http://www.sagepub.com/journalsPermissions.nav

http://usj.sagepub.com/content/44/5-6/1069.refs.html

http://usj.sagepub.com/content/44/5-6/1069.full.pdf

http://online.sagepub.com/site/sphelp/vorhelp.xhtml



Understanding Social Capital in RecentlyRestructured Urban Neighbourhoods: Two CaseStudies in Rotterdam

Reinout Kleinhans, Hugo Priemus and Godfried Engbersen

[Paper first received, November 2005; in final form, August 2006]

Summary. In the major Dutch cities, social rented housing in post-war neighbourhoods has beendemolished and largely replaced by more expensive owner-occupied and rental housing. Throughresidential mobility, these measures can trigger substantial population changes. This paper studiesresidents’ social capital in two recently restructured neighbourhoods in the city of Rotterdam. Itdistinguishes between the stayers, movers and newcomers. In a neighbourhood context, socialcapital refers to the benefit of cursory interactions, shared norms, trust and collective action ofresidents. Survey data show that social capital is not only an asset of long-term stayers, but thatin particular newcomers are relatively rich in social capital. Factors associated with higherlevels of social capital are a higher net income, presence of households with children, strongerplace attachment, higher perceived neighbourhood quality, homeownership and single-familydwellings. The expected future length of residence in the area appears of little importance forsocial capital.

1. Introduction

Urban regeneration policies are a common

phenomenon in western European countries.

Comparisons demonstrate that policy contents

and implementation differ strongly between

countries, but there are also similarities. One

is the growing importance of the concept of

social capital in the discourse of urban regen-

eration. Many policy-makers claim that urban

regeneration should not only improve the

physical quality of urban neighbourhoods,

but also the social well-being of their residents

(see for example, Flint and Kearns, 2006;

Kearns, 2004; Lelieveldt, 2004; Middleton

et al., 2005). Recently, the notion of social

capital has been introduced in the politicaldebates on urban regeneration. As Middletonand colleagues put it:

Social capital is seen as the foundation onwhich social stability and a community’sability to help itself are built; and itsabsence is thought to be a key factor inneighbourhood decline (Middleton et al.,2005, p. 1711).

To turn the tide, urban regeneration policiesoften target the housing stock of certain neigh-bourhoods. Frequent interventions are demoli-tion and upgrading of social rented housing

Urban Studies, Vol. 44, Nos. 5/6, 1069–1091, May 2007

Reinout Kleinhans is in the OTB Research Institute for Housing, Urban and Mobility Studies, Delft University of Technology, PO Box5030, 2600 GA Delft, The Netherlands. Fax: þ31 15 278 34 50. E-mail: [email protected]. Hugo Priemus is in the Faculty ofTechnology, Policy and Management, Delft University of Technology, PO Box 5015, 2600 GA Delft, The Netherlands. Fax: þ31 15278 49 33. E-mail: [email protected]. Godfried Engbersen is in the Faculty of Social Sciences, Erasmus University of Rotterdam,PO Box 1738, 3000 DR Rotterdam, The Netherlands. Fax: þ31 10 408 20 84. E-mail: [email protected]. The authors wish toacknowledge the financial assistance of the Dutch government through the Habiforum Programme Innovative Land Use and DelftUniversity of Technology through the Delft Centre for Sustainable Urban Areas.

0042-0980 Print/1360-063X Online/07/5-61069–23 # 2007 The Editors of Urban Studies

DOI: 10.1080/00420980701256047 at MEMORIAL UNIV OF NEWFOUNDLAND on August 3, 2014usj.sagepub.comDownloaded from


and new construction of owner-occupiedhousing. Urban restructuring is a commonlyused term for those measures. They result ina considerable temporary turnover of resi-dents, because significant residential mobilityout of, within and into the restructuring area isinevitable.

There is a general agreement that restruc-turing policies not only aim to improve thehousing stock and housing career opportu-nities, but also are deliberately trying topreserve or create socially mixed neighbour-hood populations (see Kearns, 2004;Kleinhans, 2004; Ostendorf et al., 2001;Tunstall, 2003). This is not an end in itself.Dutch policy-makers especially hope for animproved social liveability, a better neigh-bourhood reputation and more involvementof residents in mixed neighbourhoods(Dekker, 2005; Ministerie van VROM, 1997,2000). Attracting and retaining middle-classresidents are expected to reinforce socialnetworks of current residents and providerole models for the behaviour and aspirationsof lower-income households (Ministerie vanVROM, 1997, pp. 80–81; Uitermark, 2003).Additionally, urban restructuring policy hasshifted from a predominantly physical strat-egy to a more socially oriented and economicapproach (Kearns, 2004, pp. 4–5; Priemus,2004). Local authorities, housing associationsand care providers try to stimulate neighbour-hood involvement, common norms, mutualtrust, promoting self-help of residents andvoluntary work in community groups (Lelie-veldt, 2004; Ministerie van VROM, 2000,pp. 174–175; WRR, 2005; see also Dekker,2005, p. 15). All these issues are stronglyrelated to the concept of social capital,which Dutch policy-makers have recentlystarted to use in the urban restructuring dis-course (see WRR, 2005). It is highly likelythat social capital will become increasinglyimportant in the Dutch policy discourse, justas it did earlier in Great Britain andDenmark (Flint and Kearns, 2006, p. 33).

In the scientific literature, social capitalrefers to resources that are accessiblethrough social contacts, social networks, reci-procity, norms and trust (Bourdieu, 1986;

Coleman, 1988; Field, 2003; Putnam, 2000).In a neighbourhood context, social capitalconcerns the benefits of cursory social inter-actions, shared norms about how to treateach other and behaviour in space, trust andcollective action for a shared purpose.Several of these aspects can be recognised inpolicy-makers’ assumptions (see previousparagraph). Social capital provides a usefulanalytical perspective on the social climatein neighbourhoods after restructuring. AsPutnam states it:

Neighborhoods with high levels of socialcapital tend to be good places to raise chil-dren. In high-social-capital areas publicspaces are cleaner, people are friendlier, andthe streets are safer (Putnam, 2000. p. 307).

Although related to social cohesion, socialcapital is a different concept in severalrespects that we will briefly discuss in thispaper. This will clarify why we prefer‘social capital’ as the central analyticalconcept. Currently, we know little about resi-dents’ social capital in the context of sociallymixed neighbourhoods that have recentlyexperienced substantial population changes.The reasons are twofold. First, much researchhas concentrated on ‘traditional’ neighbourlycontacts between residents, while neglectingother social capital aspects, such as unwrittensocial norms, reciprocity and trust. Secondly,policy-makers and researchers often do notdistinguish properly between differentgroups in restructured areas. At best, they dis-tinguish between original and new residents(for example, van Beckhoven and vanKempen, 2003), following the classical studyof Elias and Scotson (1965). But what aboutresidents who moved within the same neigh-bourhood, or from directly adjacentneighbourhoods?

This contribution focuses on the socialcapital of different types of resident inrecently restructured neighbourhoods. Wewill make explicit distinctions based on theirresidential mobility patterns, their previouslocation and changes in their housing situ-ation. Thus, our empirical starting-point isthe changed neighbourhood population after

1070 REINOUT KLEINHANS ET AL.

at MEMORIAL UNIV OF NEWFOUNDLAND on August 3, 2014usj.sagepub.comDownloaded from


the implementation of the restructuringmeasures. The effects of the policy itself arebeyond the scope of the paper, because ourdata on social capital are limited to the situ-ation three years after the completion of theurban restructuring in our study areas. Unfor-tunately, no data are available for the situationpreceding the restructuring measures.

Using extensive survey data, we conductbivariate and multivariate analyses to answerthree research questions. First, what are thelevels of social capital among stayers,movers and newcomers in the neighbour-hood? Secondly, to what extent are socio-economic characteristics, neighbourhoodperceptions and housing aspects related toresidents’ social capital? Thirdly, what is therelation (if any) between social capital andresidents’ expected future length of residencein their house and neighbourhood? (seeDantas, 1988; Kleinhans, 2003). There is evi-dence for a connection between the number ofyears of residence and social capital in theneighbourhood (DiPasquale and Glaeser,1999; Saegert and Winkel, 2004). However,is social capital also influenced by residents’expectations of their future length of residence(see Middleton et al., 2005, p. 1726)? It ispossible that households score low on socialcapital because they are planning a movewithin a few years. If so, investing in goodsocial ties with neighbours and other residentsmay become less worthwhile. On the otherhand, residents may be more likely to investin social capital if they intend to stay in thearea for a long time. Our paper aims toanswer the three research questions and tocontribute to the knowledge of social capitalin neighbourhood contexts that have recentlyexperienced substantial transformation.

This paper is divided into seven sections.Next, we will describe the residential mobilityimplications of urban restructuring for differ-ent types of resident. This second sectionexplains the distinction between stayers,movers and newcomers. The third section dis-cusses theories of social capital, especially inthe context of urban neighbourhoods. Thefourth section switches to the research areas,data and methods. Fifthly, we present the

main results of bivariate and multivariate ana-lyses, followed by a discussion in the sixthsection. The final section presents the con-cluding remarks and policy implications.

2. Restructuring, Neighbourhood Dynamicsand Social Mix

In many Dutch cities, early post-war neigh-bourhoods are subject to considerableinterventions. Low-cost social rented apart-ments often dominate the housing stock inthese areas. Mainly low-income householdswith limited options rent these houses.Middle-class and higher-income householdslack attractive housing career opportunitiesin these neighbourhoods and often leavethem (Dekker, 2005; Ministerie van VROM,2000; van Kempen and Priemus, 2002;Priemus, 2004, p. 203). In 1997, the Dutchgovernment launched an ambitious restructur-ing programme to tackle the problems ofurban post-war districts (Ministerie vanVROM, 1997). Demolition, sale or upgradingof social rented housing and new constructionof more expensive owner-occupied housingshould create much more variety in thehousing stock. The neighbourhood layout,public space, services and infrastructure willbe improved at the same time. In the nextdecade, tens of thousands of households willbe directly affected (Dekker, 2005).

Urban restructuring is basically a physicalstrategy, although it is increasingly accom-panied by social and economic policymeasures. In practice, demolition and newconstruction are often so substantial thatsignificant residential mobility out of, withinand into the renewal area is inevitable. Thisrenewal-related mobility almost certainlychanges the population characteristics morefundamentally than regular residential mobi-lity patterns. The more the new and upgradeddwellings differ from the previous housingwith regard to housing type, price and tenure,the more differences in population character-istics generally arise. From that perspective,urban restructuring preserves or increases asocial mix in the neighbourhood population.

SOCIAL CAPITAL IN ROTTERDAM 1071



During the 1990s, Dutch policy-makersassumed that urban restructuring and improv-ing the quality of the housing stock were thekey to a stronger social structure and a favour-able social climate (Ministerie van VROM,1997; van Kempen and Priemus, 1999; Uiter-mark, 2003). The construction of new, moreexpensive houses, especially owner-occupied,should promote a social mix within neigh-bourhoods. This view is not exclusively aDutch policy theory. It turns up in the urbanrenewal policies of other western Europeancountries, in the US, notably in HOPE VIinitiatives (Clampet-Lundquist, 2004)—andalso in Australia (Wood, 2003). In the Nether-lands, preserving or increasing a social mix isa supposed successful strategy to combat socialsegregation and strengthen social cohesion(Ostendorf et al., 2001; Uitermark, 2003).The introduction of middle- and higher-income households was expected to alterpositively the social networks of currentresidents and provide role models inbehaviours and aspirations (van Boxtel, 2000,p. 6; Ministerie van VROM, 1997, pp. 80–81; Uitermark, 2003). Thus, the assumedconsequences of urban restructuring involveboth the ‘original’ residents and newcomers,who are supposed to earn middle or higherhousehold incomes. We will argue that thisview is too limited in its coverage of residentcategories.

As mentioned before, selective migration isone of the most pressing problems of post-warneighbourhoods. Middle- and higher-incomehouseholds often ignore these post-war neigh-bourhoods in their search for a new dwelling.However, even if these households live inthese districts, they often leave because of alack of attractive housing career opportunities(Dekker, 2005; van Kempen and Priemus,2002; Ministerie van VROM, 2000; Priemus,2004, p. 203). Precisely for this reason, thegovernment claims that restructuring shouldnot only aim at newcomers, but also targetmiddle-income households who are consider-ing a move out of areas with much socialrented housing (Ministerie van VROM,2000, pp. 176–177). Research has indeedshown that the policy succeeds in attracting

middle- and higher-income households torestructured post-war areas (Kleinhans,2005; Ministerie van VROM, 2000; Priemus,2004; van Kempen and Priemus, 2002).

Consequently, a successful restructuringpolicy may tempt wealthier residents to takeadvantage of new housing career opportunitieswithin the same area. This also applies to resi-dents who are not completely new to therestructuring site, as they live in neighbour-hoods adjacent to the area subject to restructur-ing. Therefore, we cannot study the social mixin restructured neighbourhoods in terms of asimple dichotomy of old versus new residents.The variety in moving distances, previouslocations and changes in housing situationdemands a more refined typology. Hence, wedistinguish between five resident categorieswhose mobility pattern and housing situationare directly influenced by urban restructuring(see Figure 1):

(1) Stayers who remain living in the samedwellings in the restructured area. It isquite common that only part of the neigh-bourhood is demolished. In the otherparts, restructuring measures did notrequire the stayers to move. Either theirhouses were subject to limited renovationor to no physical measures at all. Thus,many have a (far) longer length of resi-dence in the restructured neighbourhoodthan other resident categories.

(2) Movers within restructured neighbour-hoods to untouched, renovated or newlyconstructed houses. This group alsoincludes residents who experiencedforced relocation from demolished dwell-ings within the same neighbourhood.

(3) Movers from surrounding neighbour-hoods. This category includes all moversfrom adjacent neighbourhoods to therestructured area. A common finding inhousing research is that many movescover short distances (for example,Mulder and Hooimeijer, 1999).

(4) Newcomers are new residents fromanywhere outside the restructured areaand its surrounding neighbourhoods. Thenewcomers mainly moved to the newly




constructed houses, but also to the orig-inal or the renovated houses.

(5) Forced movers out of restructured areas:residents who are forced to move to adifferent neighbourhood, due to demoli-tion or upgrading of their dwelling. Theissue of forced relocation is beyond thescope of this paper, because our focus ison social capital in recently restructuredneighbourhoods. However, the issue isstudied in depth in other papers(Allen, 2000; Clampet-Lundquist, 2004;Ekstrom, 1994; Fried, 1967; Gans, 1991;Goetz, 2002; Kleinhans, 2003; Popp,1976).

As mentioned in the introduction, our researchdoes not allow a ‘pure’ ex-post evaluation thatuses data of the situation before and after theintervention. Still, our resident typology canbe related to the extent to which the currentpopulation characteristics are the result ofurban restructuring. Subsequently, we cananalyse current levels of social capital andcompare the stayers, movers and newcomers.However, the application of the socialcapital concept in a neighbourhood contexthas its pitfalls. In the next section, we there-fore explore the theoretical backgrounds ofsocial capital.

3. Social Capital in a NeighbourhoodContext

3.1 Theories of Social Capital

Social capital has received much internationalattention through the works of Bourdieu(1986), Coleman (1988) and Putnam (1993,2000). However, it is by no means a novelty:

The term social capital itself turns out tohave been independently invented at leastsix times over the 20th century, each timeto call attention to the ways in which ourlives are made more productive by socialties (Putnam, 2000, p. 19).

Definitions differ per author. Generally, socialcapital refers to resources that are accessiblethrough social interactions and social net-works, reciprocity, norms and mutual trust(Bourdieu, 1986; Coleman, 1988; Fine,2001; Portes, 1998; Putnam, 1993, 2000; forcomprehensive overviews of the literature onsocial capital, see Field, 2003; Fine, 2001;Halpern, 2005; and Kearns, 2004).

The usefulness of social capital as ananalytical concept has been questioned in thescientific debate (see Middleton et al., 2005,pp. 1713–1717 for an overview). Basically,however, the concept of social capital is

Figure 1. A graphical model of the five categories of residents.




rather straightforward. By making connectionswith one another, and maintaining these con-tacts over time, people are able to worktogether. They are able to achieve things thatthey either could not achieve by themselves,or only with difficulty and at high costs. Tothe extent that social interactions and networksconstitute a resource, they form a kind ofcapital (Field, 2003, p. 1). However:

it is important to distinguish the resourcesthemselves from the ability to obtain themby virtue of membership in differentsocial structures, a distinction explicit inBourdieu but obscured in Coleman(Portes, 1998, p. 5).

Portes defines social capital as the ability tomobilise resources from a social network.Thus, an individual must be connected toothers to reap social capital benefits. Apartfrom definition issues, many authors observea distinction between inward-looking andoutward-looking social capital (see Table 1).

While a few authors have added a thirddimension—i.e. linking capital (for example,Halpern, 2005; Woolcock, 1998)—the dis-tinction between bonding and bridgingcapital has received most attention. Bondingcapital is a resource created in the strongsocial ties between individual people—i.e.certain family members, close friends andmembers of certain ethnic groups. Strongties are a major source of emotional andmaterial support (bonding capital). This typeof capital can be very important within poor

and excluded communities (Kearns, 2004).The social networks that produce bondingcapital can be so strong that they exclude out-siders from the network and impose suffocat-ing norms on the group members (Briggs,1998; Portes, 1998). This is known as thedark side of social capital (Portes andLandolt, 1996; Putnam, 2000).

Bridging capital is hidden in the weak, lessdense, cross-cutting social ties betweenheterogeneous individuals such as friends ofyour friends, indirect acquaintances orcertain colleagues from your work. Thisform of capital helps people to ‘get ahead’through access to opportunities and resourcesin other social circles than their own. Thus, itcontains a different type of resource frombonding capital. A classic example of brid-ging capital is information about job opportu-nities, passed on between loosely connectedpeople through a common acquaintance. Theweak ties concept originated from the researchof Granovetter.

Whatever is being diffused can reach alarge number of people and traversegreater social distance when passedthrough weak ties rather than strong ties(Granovetter, 1973, p. 1371).

Thus, bonding and bridging capital have theirown specific merits and drawbacks. But theyare not ‘either–’or’ categories into whichsocial networks can be neatly divided, but‘more or less’ dimensions along which wecan compare different forms of social capital

Table 1. Two types of social capital

Socialcapital

Granovetter(1973) Henning

and Lieberg(1996) Briggs (1998)

Woolcock(1998)

Gittell and Vidal(1998) Putnam

(2000) Lin (2001)

Internally Strong ties Social support(to get by)

Integration(group)

Bonding capital Expressive action(homophilousties)

Externally Weak ties(bridges)

Social leverage(to getahead)

Linkage(betweengroups)

Bridging capital Instrumentalaction(heterophilousties)




(Putnam, 2000, p. 23). If Dutch policy-makerstry to stimulate shared social norms, involve-ment and collective action between residents,they appear to aim mainly at weak ties andbridging capital. However, the application ofthese network terms is problematic in a neigh-bourhood context. We will next discuss this indetail.

3.2 Social Capital in Neighbourhoods

Because social capital and weak ties are basi-cally network concepts, studying social capitalin neighbourhoods poses several problems.1

Most important is that neighbourhoods and net-works are completely different entities thatalmost never converge (see Wellman et al.,1988). ‘Neighbourhood’ is a socio-spatial orimagined unit with a specific, but a limited,social significance for its residents. It is onlyone of the many contexts in which people estab-lish and maintain their social networks. Thus,neighbours and other residents usually formjust a small part of residents’ social networks(Bridge, 2002, p. 25; Fischer, 1982, p. 41;Henning and Lieberg, 1996).

However, the neighbourhood is a contextthat residents choose or are forced to live in.Therefore, we are interested in cursory, every-day social interactions between residents thatmay produce social capital without necess-arily being a member of each other’snetwork. These cursory ties may developinto strong ties (bonding), but they usuallyremain of a weak nature and of “a shifting,moving, fluid character” (Lofland, 1985,p, 118). In her book A World of Strangers,Lofland studied social interactions in publicspace, characterised by limited verbal com-munication and a short duration. WhileLofland emphasises evasive behaviour, wewill argue that cursory social interactionsmay have a positive social capital value.

Like pennies dropped in a cookie jar, eachof these encounters is a tiny investment ofsocial capital (Putnam, 2000, p. 93).

Neighbourhood residents ‘accidentally’ runinto personal encounters in staircases, on thestreet, on squares, in playgrounds and in

neighbourhood facilities such as shops andcommunity centres. To a certain degree,there is a form of mutual dependency. Thisdependency is hidden in the extent to whichresidents live peacefully alongside eachother, succeed in maintaining commonnorms and trust, and co-operate successfullyif a shared neighbourhood interest is atstake. In a negative sense, the dependencybetween residents is felt clearly if nuisanceoccurs. However, the benefits of sharednorms, trust and collective action are aresource from cursory, everyday social inter-actions. Consequently, these benefits areforms of social capital.

Cursory social interactions can yield publicfamiliarity. Public familiarity implies thatresidents get sufficient information fromeveryday interactions to recognise and ‘cat-egorise’ other people (Fischer, 1982, pp.60–61; see also Blokland, 2003, pp. 90–93).Public familiarity can result in social capitalin the sense of a favourable social climate,but also in more tangible forms of socialcapital. We will give some examples toclarify our argument. We first refer to workof Henning and Lieberg (1996), who studiedthe role of weak ties between residents.They define weak ties as the “unpretentiouseveryday contacts in the neighbourhood”(p. 6). These contacts range from a noddingacquaintance to modest levels of practicalhelp. The number of weak ties outnumberedthe strong ties. Weak ties not only appearedto be significant for support, but also for afeeling of home and security (Henning andLieberg, 1996; Briggs, 1998; Skjaevelandand Garling, 1996; see also Crawford, 2006).Forrest and Kearns argue that:

the less robust and less deep-rooted areneighbourhood networks, the more stableand conflict-free may be the social orderin which they sit (Forrest and Kearns,2001, p. 2134).

According to Bridge (2002, p. 15), what wecan reasonably expect from other residents isneighbourliness. This is the exchange ofsmall services or support in an emergencyagainst a background of routine convivial




exchanges, such as greetings and brief chatsover the garden fence or in the street.

A second element of social capital concernssocial norms. In a neighbourhood setting,norms are unwritten social rules and opinionswith regard to social interactions with otherresidents and behaviour in public spaces.Social capital then consists of benefits ofshared norms and social control, such as nui-sance that fails to occur, agreements on howto use scare parking space and parents alsokeeping an eye on playing children otherthan their own (see Foley and Edwards,1999; Putnam, 2000). Related is the conceptof collective efficacy, defined as social cohe-sion among neighbours combined with theirwillingness to intervene on behalf of thecommon good (Sampson et al., 1997,p. 918). Sampson and colleagues showedthat collective efficacy is negatively associ-ated with variations in violent crime in neigh-bourhoods. Residents’ willingness tointervene in unpleasant situations partlydepends on the quality of social interactionsand mutual trust (Sampson et al., 1997,p. 919; Coleman, 1990; Duncan et al.,2003). Social capital theory claims that effec-tive enforcement of norms is only possible if asocial structure has closure (Coleman, 1988,pp. S105–S107). Closure refers to the extentto which different actors in a social settingare interconnected—i.e. know each other. Ina neighbourhood, this would mean thatresidents must know each other if theywant to exercise social control. However,Bellair (1997) has suggested that themere presence of social interactions is suffi-cient for a basic level for social control.Moreover, certain explicitly agreed normscan be enforced top–down by landlords.They can also stimulate initiatives of residentswho want to draw up basic norms fortheir apartment buildings. This ‘codification’may simplify residents’ efforts of normenforcement.

Trust, a third component of social capital, isa complex issue.

The causal arrows among civic involve-ment, reciprocity, honesty, and social trust

are as tangled as well-tossed spaghetti(Putnam, 2000, p. 137).

A basic level of trust is a condition for socialinteraction, support and reciprocity. Trustmay also develop as a positive consequenceof interactions and mutual support (Brehmand Rahn, 1997). In a neighbourhoodcontext, trust refers mainly to predictabilityof residents’ behaviour. A deteriorating neigh-bourhood poses threats to this predictabilityand social interactions between residents(Fukuyama, 1995, p. 26; Lelieveldt, 2004;Ross et al., 2001). However, an improvingneighbourhood may have beneficial effectsfor trust levels. Residents may perceiveinvestments in the physical infrastructure asa sign of public interest in their neighbour-hood, raising their optimism and trust in itsfuture (see also Flint and Kearns, 2006, p. 45).

In sum, we have described how socialcapital can be analysed in a neighbourhoodcontext. While strong ties in a neighbourhoodcan produce bonding social capital, it seemsthat weak ties—i.e. casual and cursory con-nections between residents—are far morelikely to occur. These connections canproduce a variety of resources, all supportinga favourable social climate. Therefore, socialcapital has both an individual and a collectivedimension. The resources accrue to both the‘groups’ of residents involved in cursory con-nections, as well as individual residents. Butthere is no such thing as the social capital ofa neighbourhood, which is not a socialentity. Streets and building blocks are farmore important levels for social interactionthan the neighbourhood level, especiallyin mixed-tenure neighbourhoods (see forexample, Jupp, 1999). We designed oursurvey to match this line of reasoning. More-over, we acknowledge the results of recentDutch research, which found that social mixas a result of urban restructuring has notimproved social cohesion, but has resulted insocial divisions at the neighbourhood level(see van Beckhoven and van Kempen, 2003;Dekker and Bolt, 2005). However, here weuse a different analytical perspective, both inspatial terms (interactions between relatively




small ‘groups’ of people at the micro scale ofstreets and building blocks) and in terms of thecentral concept: social capital.

That brings us to some important differ-ences between social cohesion and socialcapital. Whereas social capital refers toresources accessible through social networks,norms and trust, social cohesion commonlydenotes the networks, values, norms and soli-darity themselves. Social cohesion often con-jures up the notion of intensive relations insocial networks, while social capital paysmore attention to the added value ofmany-branched networks and weak ties.Additionally, social capital is, by definition,limited to interactions between people,excluding relations between people andplaces. Interestingly, several authors considersocial capital as a dimension of social cohe-sion, with the other dimensions beingcommon values and civic culture, socialorder, solidarity and place attachment/iden-tity (for example, Forrest and Kearns, 2001,Dekker and Bolt, 2005). Restricting ourselvesto the concept of social capital, we can deducethat place attachment is not a part of socialcapital. However, we do not exclude a con-nection between these concepts beforehand,which is why place attachment is introducedas an independent variable (see next section).

Finally, we discuss the linkage between resi-dential mobility and social capital. As mentionedearlier, there is evidence for a connectionbetween years of residence and (preparednessto contribute to) social capital in the neighbour-hood (DiPasquale and Glaeser, 1999; Saegertand Winkel, 2004). Nevertheless, residents’expectations of their future length of stay in theneighbourhood may also be associated withsocial capital. A tendency to move in the nearfuture may have a negative impact on thesocial capital of households. Whether they areplanning a move, is reflected by their expectedlength of residence (see Dantas, 1988; Klein-hans, 2003). Research has shown that residents,who claim to move within a few years, canusually indicate the main triggers of their inten-tions (Mulder and Hooimeijer, 1999). Thisfinding was replicated in the research reportedin this paper (Kleinhans, 2005).

4. Data and Methods

4.1 Data Collection

Understanding social capital of stayers, moversand newcomers in restructured neighbourhoodsrequires thorough empirical research. Ourfinancial resources enabled extensive fieldworkand data analysis in only two peripheral post-war neighbourhoods in the city of Rotterdam:De Horsten and Hoogvliet Northwest. Bothneighbourhoods were built during a period ofsevere housing shortage as a result of theSecond World War. The area of Hoogvlietwas also meant for housing employees of thepetrochemical industry nearby. The areaswere dominated by multifamily apartmentbuildings in the social rented sector. Duringthe 1990s, extensive urban restructuring trans-formed the housing stock of De Horsten andHoogvliet Northwest. Nowadays, both neigh-bourhoods consist of almost 1000 dwellingsof different forms, tenures, prices and quality.We distributed 1941 written questionnairesamong all households in both study areas. Sub-sequently, we collected the questionnaires in apersonal door-to-door campaign. This yieldeda response of 917 useable questionnaires—i.e.47 per cent—almost equally spread betweenthe areas. Then, we acquired neighbourhoodcensus data, such as household composition,age, ethnic background and tenure. These datawere compared with the equivalent survey vari-ables. This analysis (not printed here) showedthat the response is a representative sample ofthe population in both areas. Several questionsin the questionnaire enabled categorisation ofrespondents (see Table 2).

A striking difference between the researchareas is the relative size of the categories. InHoogvliet Northwest, stayers are the largestcategory and also form a much bigger shareof the population than in De Horsten. On theother hand, the proportion of newcomers ismuch higher in De Horsten than in HoogvlietNorthwest. The sheer size and nature of therestructuring measures determine these differ-ences. In De Horsten, about 70 per cent of thehousing stock has been demolished and reno-vated. For Hoogvliet Northwest, this figureamounts to 40 per cent. Nevertheless, the




proportion of movers within the neighbourhoodis the same in the response. The newcomers inboth areas are mainly from other districtsin Rotterdam, as well as other municipalities.

4.2 Measures

From the previous section, it is clear thatsocial capital is a multidimensional concept(Foley and Edwards, 1999; Fine, 2001;Narayan and Cassidy, 2001; Putnam, 2000).The survey contains 22 indicator variables ofsocial capital (see the Appendix). The vari-ables both reflect the nature of the specifictype of social capital and the way in which itcan be ‘accessed’ by the respondents. Asmentioned earlier, all variables on social inter-actions, norms and trust are designed in a wayto indicate cursory connections, but not toexclude possible strong ties. Most variablesare measured on a five-point Likert scale.Yet, it makes no sense using each item as adependent variable. We need a compositemeasure that simultaneously includes aspectsof social interactions, norms and trust. There-fore, we combined all variables in a socialcapital index (see Putnam, 2000, p. 291).Cronbach’s a-coefficient of this index is0.75. We also constructed several indices ofcomponents of social capital. A principalcomponents analysis (see Appendix) has indi-cated three relevant components: social inter-actions and the resulting public familiarity;norms and trust; and associational activity.However, the two separate indices of normsand trust, and associational activity haveCronbach’s a-values just below 0.7 (seeAppendix), which is the widely accepted

social science cut-off for the a-value.However, as we concentrate on the analysisof the social capital index, there is no validityproblem.

The multivariate analysis includes severalpotential predictors of social capital. First, weuse expected length of residence (ELR), neigh-bourhood and resident category, as defined inTable 2. ELR is a dummy variable, indicatingan expectation to move in less than five yearsor not, measured at the moment of answeringthe question. Secondly, to capture respondents’socioeconomic characteristics, we include age(years), household composition (householdswith or without children), labour market pos-ition (paid employment or otherwise), nethousehold income per month (lower versusmiddle and higher income) and ethnic back-ground (native Dutch or ethnic minority). InDutch statistics, a person belongs to an ethnicminority if at least one of his parents wasborn abroad, regardless of his own country ofbirth.2 Finally, data on educational levels ofrespondents were not available.

Measures of housing and neighbourhoodperception are also included. The housing vari-ables include satisfaction with the currentdwelling (a scale ranging from 1 ¼ very unsa-tisfied to 5 ¼ very satisfied), tenure (social orprivate rented versus owner-occupied) anddwelling type (single- or multifamily). Theremaining variables are place attachment andperceived neighbourhood quality. These twovariables need more clarification.

Place attachment involves dynamic, butenduring, positive bonds between residentsand their physical settings (Brown and Perkins,1992). In our research, place attachment is an

Table 2. Categories of residents in De Horsten and Hoogvliet Northwest

Category

De Horsten Hoogvliet Northwest

N Percentage N Percentage

Stayers 42 9.0 199 44.4Movers within the neighbourhood 63 13.4 58 12.9Movers from surrounding neighbourhoods 136 29.0 94 21.0Newcomers 219 46.7 96 21.4Missing (unknown) 9 1.9 1 0.2

Total (N ¼ 917) 469 100.0 448 100.0




index based on nine questionnaire items and con-tains the mean score per respondent for all items.These items are statements reflecting the extentto which respondents appreciate living in theneighbourhood, feel proud of it, feel at home inthe area, perceive accessible moving opportu-nities and feel safe on the streets at night(Brown et al., 2003; Burns et al., 2001; Forrestand Kearns, 2001; Perkins and Long, 2002).Each item is measured on a five-point Likertscale. Scales with reversed meanings wererecoded accordingly. The scores of the indexrange between 1 and 5 (Cronbach’s a ¼ 0.84).

Perceived neighbourhood quality is ameasure of residents’ perceptions of the phys-ical quality of their immediate living environ-ment. Here, we developed an index that isconstructed similarly to the place attachmentindex. Perceived neighbourhood quality con-sists of five items measuring how often vandal-ism, graffiti on buildings, litter and dog dirt onthe streets, nuisance from other residents andunsafety on the streets occur, according to therespondent (see Brown et al., 2003; Ellawayet al., 2001; Parkes et al., 2002). Each item ismeasured on a four-point scale (1 ¼ oftenoccurs here, to 4 ¼ never occurs). Again, thecoding of certain items was reversed to takenegative statements into account. Scores ofthe perceived neighbourhood quality indexrange between 1 and 4 (Cronbach’s a ¼ 0.80).

5. Results

As a starting-point, we calculated the averagesocial capital index (SCI) score for each of the

resident categories in our research areas (seeTable 3). This yields notable results. First,the groups in both neighbourhoods differsignificantly in the average SCI score. In DeHorsten, stayers have a much lower level ofsocial capital than the movers and newcomers.By contrast, the stayers score highest of allgroups in Hoogvliet Northwest. Secondly,the newcomers in both areas have a relativelyhigh level of social capital. If length of resi-dence is a strong predictor of social capital,the newcomers would score much lowerthan stayers. Finally, the total average SCIscore does not differ significantly betweenthe research areas.3

We used linear regression analysis to estab-lish the predictors of residents’ social capital.The regression analysis of the social capitalindex consists of three models, of whichonly the final model is depicted in Table 4(second column). The final model includesthe expected length of residence (ELR), theresident classification, the neighbourhood des-ignation, interaction terms for joint effects ofresident category and neighbourhood, socio-economic characteristics, dwelling satisfac-tion, place attachment, tenure, dwelling typeand perceived neighbourhood quality.4

Table 4 confirms that the movers from thesurrounding neighbourhoods score lower onsocial capital than newcomers. To someextent, this goes against the expectations ofpolicy-makers. They would expect higherlevels of social capital with residents whomoved only a relatively short distance—i.e.from adjacent neighbourhoods. Table 4 also

Table 3. The social capital index: mean scores per resident category (N ¼ 871)

Areas Stayers

Movers withinrestructured

neighbourhoods

Movers fromsurrounding

neighbourhoods NewcomersAverageper area

De Horsten 2.27 2.64 2.68 2.71 2.65(SD) (0.39) (0.41) (0.32) (0.34) (0.37)Hoogvliet Northwest 2.73 2.68 2.61 2.68 2.68(SD) (0.32) (0.27) (0.34) (0.33) (0.32)

Notes: Social capital index: all respondents with more than five missing values for variables in the index are excluded. The

higher the index score, the higher the average level of social capital of the resident category (index range: 1–5). De Horsten:

ANOVA sum of squares between groups ¼ 6.62; df ¼ 3; F ¼ 18.06; p , 0.001. Hoogvliet: ANOVA sum of squares between

groups ¼ 0.87; df ¼ 3; F ¼ 2.83; p , 0.05.




Table 4. Predictors of residents’ social capital and of separate social capital components (C1–C3)(N ¼ 781)

Dependent variables

Social capitalindex (final

model)C1 Social

interactions indexC2 Norms and

trust indexC3 Associational

activities

B SE B SE B SE B SE

Category of residentsStayers 0.04 0.04 0.05 0.08 0.09 0.07 0.14 0.26Movers within the

neighbourhood20.07 0.05 20.11 0.09 20.09 0.09 20.14 0.26

Movers fromsurroundingneighbourhoods

20.10� 0.04 20.17� 0.08 20.12 0.08 20.21 0.22

Newcomers (referencecategory)

0 0 0 0

Neighbourhood(0 ¼ Hoogvliet;1 ¼ Horsten)

0.02 0.04 20.05 0.06 0.03 0.07 20.03 0.20

Interaction: category� neighbourhoodStayers 20.21�� 0.07 20.36�� 0.12 20.23 0.12 – –Movers within the

neighbourhood0.08 0.07 0.12 0.12 0.15 0.12 – –

Movers fromsurroundingneighbourhoods

0.12� 0.06 0.21� 0.09 0.14 0.09 – –

Newcomers (referencecategory)

0 – –

Expected length ofresidence (ELR)(0 ¼ more than fiveyears; 1 ¼ less thanfive years)

0.05 0.03 0.01 0.05 0.06 0.05 0.24 0.23

Age (in years) 0.00 0.00 0.00 0.00 0.00 0.00 0.02�� 0.01Ethnicity (0 ¼ ethnic

minority; 1 ¼ nativeDutch)

0.01 0.02 0.16�� 0.04 20.17�� 0.04 0.10 0.18

Household composition(0 ¼ HH withoutchildren, 1 ¼ HHwith children)

0.09�� 0.03 0.12�� 0.05 0.12�� 0.04 0.33 0.20

Labour market position(0 ¼ unemployed,retired; 1 ¼ paidemployment)

20.06� 0.03 20.13� 0.05 20.01 0.05 0.39 0.22

Net household incomeper month (0 ¼ lowincome, ,E1.500;1 ¼ middle or higherincome, .E1.500)

0.07�� 0.03 0.14�� 0.05 20.07 0.04 0.36 0.19

Satisfaction with thecurrent dwelling

0.03 0.02 0.06� 0.03 0.05 0.03 0.11 0.11

Place attachment(index)

0.22�� 0.02 0.30�� 0.04 0.23�� 0.04 0.56�� 0.17

(Table continued)




shows the substantial difference in socialcapital scores between the stayers in bothareas. In the next section, we provide anexplanation for this.

Apart from the final regression model of thesocial capital index, Table 4 also depicts separ-ate regression analyses of the three socialcapital components (see the Appendix). Thisenables us to study how the results on thesocial capital components differ from theresults for the overall level of social capital. Itappears that the social capital index model(R2 ¼ 0.39) has more explanatory power thanthe models explaining the social interactionsand norms and trust indices (R 2 ¼ 0.35 and0.28 respectively). The fit of the associationalactivities model is not very satisfying (Nagelk-erke R2 ¼ 0.11). As mentioned earlier, wefocus our attention on the social capital index,in the second column of Table 4.

The regression model of the social capitalindex shows no significant relationshipbetween social capital and the ELR, whilethere was a highly significant connection inthe two preceding models (not shown here).The positive association between socialcapital and ELR disappeared after controllingfor housing characteristics, dwelling satisfac-tion, place attachment and perceived neigh-bourhood quality.

Both age and ethnic background have nosignificant association with the overall levelof social capital. However, age is a significantpredictor of associational activities—i.e.active membership of a local association,volunteering and collective action with resi-dents (see Middleton et al., 2005). Moreover,ethnic background has a significant relation-ship with both the social interactions indexand the norms and trust index.

Table 4. Continued

Dependent variables

Social capitalindex (final

model)C1 Social

interactions indexC2 Norms and

trust indexC3 Associational

activities

B SE B SE B SE B SE

Tenure (0 ¼ rented;1 ¼ owner-occupation)

0.07� 0.03 0.08 0.05 0.04 0.05 0.46� 0.21

Dwelling type(0 ¼ single-familyhome; 1 ¼ multi-family dwelling)

20.08�� 0.03 20.11� 0.05 20.06 0.05 0.11 0.22

Perceivedneighbourhoodquality (index)

0.08�� 0.02 0.11�� 0.03 0.17�� 0.03 20.39� 0.14

Constant 1.58�� 0.10 1.32�� 0.18 1.67�� 0.17 22.88�� 0.76F 23.97 19.64 14.71 – –Df 18 18 18 15Significance 0.000 0.000 0.000 0.000R2 0.39 0.35 0.28 0.11

(Nagelkerke R2)

Notes: Linear regression for the indices social capital, social interactions, and norms and trust. Logistic regression for asso-

ciational activity: 0 ¼ no participation; 1 ¼ participation (i.e. active membership of a local association and/or volunteering

and/or collective action). Significance levels: �p , 0.05, ��p , 0.01, ��p , 0.001 (two-tailed). All respondents with more

than five missing values for variables in the social capital index are excluded from the analyses. This step decreases the

number of incomplete index values for the three other indexes to 6 per cent or less. The linear regression models meet the

requirements of multiple regression: linearity of relationships and homoscedasticity (tests of these assumptions can be

requested from the first author).




Households with children at home scorehigher on social capital. Labour market pos-ition—i.e. having a paid job—has a negativeeffect on both social capital and the perceptionof the social interactions. Of all the socioeco-nomic characteristics, household income isthe strongest predictor. Middle- and higher-income households have a higher level ofsocial capital than lower-income households.The analysis of the social interactions indexrepeats this finding, but not the other twocomponents.

Finally, we analyse the effect of housingand neighbourhood characteristics. Theinclusion of the corresponding variables inthe final SCI model has resulted in a substan-tial improvement of the explanatory power.5

Dwelling satisfaction has no relation withsocial capital. However, the strength of theconnection between place attachment andsocial capital is remarkable (b ¼ 0.39,p , 0.001). It reappears in the analysis ofeach of the separate social capital com-ponents. The same applies to the perceivedneighbourhood quality, except for its negativeassociation with associational activities.

Two other significant factors are dwellingtype and tenure. First, owner-occupiers scorehigher on social capital than renters. Sec-ondly, living in a single-family dwelling isassociated with higher levels of socialcapital than living in multistorey apartments.In conclusion, not only socioeconomicfactors but also housing and neighbourhoodcharacteristics play an important role inexplaining residents’ level of social capital.

6. Discussion

We have applied the concept of social capitalin the context of two recently restructuredneighbourhoods that have experienced sub-stantial residential and social instability.Thus, the results are only valid for our casestudies and not necessarily for the generalDutch situation. Our multivariate analysisshows a variety of factors explaining thesocial capital scores. Several socioeconomiccharacteristics are important for the level ofsocial capital. Households with children at

home clearly stand out in comparison withhouseholds without children. Obviously,parents meet other parents by means of theirchildren—for example in the playground orin the schoolyard. This is a common way toget to know other residents (see Forrestand Kearns, 1999; Saegert and Winkel,2004). These interactions can increase publicfamiliarity between residents. If experiencedas positively, they are likely to producesocial capital in various forms (see section3.2).

Although ethnic background is not signifi-cantly related to the overall level of socialcapital, it is associated with two separate com-ponents. Native Dutch respondents are morepositive about social interactions and publicfamiliarity in their neighbourhood thanethnic minorities. On the other hand, ethnicminorities appear significantly more positiveabout levels of norm conformity and trustbetween residents than native Dutchmen. Wecan only speculate about this contradiction.Native Dutch respondents may have moreopportunities to engage in social interactionswith other (native) residents. Simultaneously,they may have less trust in other residents,especially in those with other ethnic back-grounds. Possibly, the prejudices of nativeDutch residents against ethnic minoritiesare stronger than the other way round (seeDekker and Bolt, 2005, p. 2467). In thesocial capital index, these contrasting associ-ations seem to cancel each other out.

Closer inspection of our household incomedata (not shown) reveals that we are dealingpredominantly with middle-income householdsand relatively few high-income households.6

Both middle- and higher-income householdshave a significantly higher level of socialcapital than low-income households (see alsoButler and Robson, 2001; Drukker et al.,2005; Saegert and Winkel, 2004). Presumably,this is a joint effect of economic and culturalcapital—i.e. the level of education and skills.Bourdieu (1986) conceptually describes theexchange processes between economic, socialand cultural capital (see also Piachaud, 2002;Silva and Edwards, 2004). Although we lackdata on education, we hypothesise that such




an exchange process explains the strong posi-tive connection between the householdincome and level of social capital. More specifi-cally, Middleton et al. (2005) claim that thepresence of (bridging) social capital is a conse-quence of social and economic well-being, nota cause of it. They write that membership ofmany organisations, such as sport clubs,requires wealth in order to invest time andmoney in participation (p. 1731, 1734). Thissuggests that a potential causal arrow pointsfrom economic capital to social capital.

Another explanation for the correlationbetween income and social capital level isthe time of arrival in the area. Most of themiddle- and higher-income householdsmoved into the restructured neighbourhoodsin a limited period of time after the com-pletion of the new dwellings. Consequently,their length of residence is very similar.They experienced a joint new start in theneighbourhood. Research into new estatesshows a relatively high level of social inter-action in the first years of the estate, andthose interaction levels tend to diminish after-wards (for example, Reijndorp et al., 1998;Jupp, 1999). This finding seems to apply toour study-area neighbourhoods as well (seealso Dekker and Bolt, 2005, p. 2461). Themiddle- and higher-income households arerepresented most among the newcomers.This partly explains why stayers and moverswithin the neighbourhood do not surpass thenewcomers’ social capital level.

Somewhat paradoxical is our empiricalfinding that paid employees have relativelyless neighbourhood-based social capital thanresidents who are retired, unemployed orotherwise fall outside the labour market. Asthe household income is closely linked topaid employment, we expected the labourmarket position to have a positive effect.The underlying cause is a strong heterogeneityof respondents within the categories paidemployment, retired and unemployed. Forexample, significantly more pensioners fromthe middle-income-group are active volun-teers than respondents in paid employmentfrom the same income-group. Furthermore,people with jobs are more often and longer

away from their neighbourhood. Thus, theyseem to be part of further-reaching networksand their opportunities to invest in neighbour-hood-based social capital are more limited,both in terms of available time and diversityof presence during the day. The effect of theposition in the labour market is, however,overshadowed by the effect of income andhousehold composition on social capital.

Place attachment has a remarkably strongassociation with the level of social capital.In other words, residents expressing a higherlevel of place attachment also report higherlevels of (access to) social capital. Severalauthors have hinted at such a relationbetween place attachment and social capital(Burns et al., 2001, p. 7; DiPasquale andGlaeser, 1999; Forrest and Kearns, 2001,p. 2140; Perkins and Long, 2002). We con-sidered the possibility that residents feeltrapped in the neighbourhood, and, simul-taneously report high levels of place attach-ment (see Vale, 1997). However, our placeattachment index takes into account residentswho prefer to move, but simply lack theresources and opportunities to act accord-ingly. A stronger place attachment is likelyto raise residents’ willingness to join infavourable social interactions that createsocial capital. However, place attachmentmay result partly from the availability andreproduction of social capital among resi-dents. Due to the cross-sectional nature ofthe data, we cannot trace a potentially causalrelation between place attachment and socialcapital.

Perceived neighbourhood quality is alsocorrelated with the level of social capital.However, the positive connection of this indi-cator with the social interactions and normsand trust indices contrasts with the negativeconnection with associational activities.Thus, a more negative neighbourhoodopinion raises the chances of participation inassociations, volunteering or collectiveaction with neighbours. Such a (plausible)relationship would mean that a lower per-ceived neighbourhood quality stimulates resi-dents to initiate efforts, whether jointly or not,to improve the neighbourhood attractiveness.




Dwelling characteristics also matter forsocial capital. Owner-occupiers in the restruc-tured neighbourhoods enjoy higher levels ofsocial capital than the renters, whetherprivate or social renters. This finding fitsneatly in a research tradition that points tothe beneficial effects of homeownership forboth the owner-occupiers and the neighbour-hood (see Campbell and Lee, 1992; Davidsonand Cotter, 1986; DiPasquale and Glaeser,1999; Elsinga and Hoekstra, 2004; Temkinand Rohe, 1998). Here, owner-occupiers par-ticipate significantly more often than rentersin associational activities and volunteering,both in and outside their neighbourhood.They have, therefore, more access to socialnetworks potentially rich in social capital.The home is also an asset and its value isclosely tied to the attractiveness of the com-munity (Middleton et al., 2005, p. 1728).Motivated by protecting their investment,homeowners may be more likely to organisethemselves. They may be able to endorseunwritten codes of conduct more easily thantenants.

We ascertained earlier that length of resi-dence in a neighbourhood is often seen as apositive co-determinant of residents’ socialcapital. A separate multivariate analysis (notshown here) demonstrated that length of resi-dence did not emerge as a distinctive factor.The main reason is the recent arrival ofmovers from adjacent neighbourhoods andthe newcomers. We therefore used expectedlength of residence in multivariate analy-sis—i.e. the time that residents expect tostay in their current dwelling (Dantas, 1988;Kleinhans, 2003). The assumed effect wasthe same as the effect of length of residencethat had been discounted. Nevertheless,Table 4 revealed that the relationshipbetween expected length of residence andsocial capital is not significant. This holdstrue if, besides the social characteristics ofthe residents, account is taken of dwelling sat-isfaction, place attachment and the dwellingcharacteristics. This is comparable with awell-known conclusion of Kasarda and Jano-witz (1974). They established that residentswant to leave the local community if it fails

to meet their aspirations, despite a strongneighbourhood attachment and intensivelocal participation (p. 329).

Residents from multifamily dwellings reportsignificantly lower levels of social capital thanrespondents living in single-family dwellings.This may be partly explained by a combinedeffect of tenure, a higher tendency to moveand the actual length of residence of peopleliving in apartment blocks or single-familyhomes. This combination is different in eachneighbourhood. In De Horsten, all single-family homes are new owner-occupiedproperties. All old dwellings are socialrented apartments, from which significantlymore residents reported a tendency to move.In Hoogvliet Northwest, the relationshipsbetween dwelling type and tenure were lessstraightforward. Even so, the effect of thetype of dwelling on social capital is still sig-nificant if resident category, neighbourhood,social characteristics, expected length ofresidence, satisfaction with the dwelling,place attachment and tenure are held constant.Another explanation is that it is harder tocreate and maintain pleasant social inter-actions and shared norms in old apartmentblocks with a high occupancy turnover thanin terraced dwellings with a very low occu-pancy turnover. Moreover, evidence fromother studies suggests that people living inflats are less likely to chat to a large numberof neighbours (Middleton et al., 2005,p. 1726). This probably decreases opportu-nities for the creation of social capital.

Finally, we return to the finding that stayersin De Horsten display much lower socialcapital scores than the stayers in HoogvlietNorthwest. In De Horsten, the stayers exclu-sively live in old, social rented multifamilydwellings that contrast negatively with otherparts of De Horsten. These stayers predomi-nantly earn low incomes and are less attachedto the neighbourhood. Additionally, their sat-isfaction with their housing and neighbour-hood situation strongly lags behind those ofother groups. Their building blocks arecharacterised by higher levels of occupancyturnover than other parts of the neighbour-hood. Consequently, the stayers in De




Horsten are confronted with frequently chan-ging neighbours. On the other hand, manystayers in Hoogvliet Northwest live inowner-occupied single-family dwellings andearn modest but not low incomes. They arerelatively often native Dutch empty-nestersthat have been living for a long time in astable part of Hoogvliet Northwest. Further-more, their place attachment and satisfactionwith their housing and neighbourhood situ-ation are comparable with those of themovers and newcomers in Hoogvliet North-west (Kleinhans, 2005). In sum, both struc-tural and attitudinal characteristics explainthe differences between stayers in DeHorsten and Hoogvliet Northwest.

7. Concluding Remarks

This paper has focused on the social capital offour different resident categories in two Dutchrestructured post-war neighbourhoods. Withinthis context, we have defined social capital asthe benefit of cursory interactions, sharednorms, trust and collective action of residents.In our two restructured neighbourhoods,social capital is hardly an asset at the neigh-bourhood level, but usually appears at amore local level: in building blocks, streets,parks, playgrounds and over garden fences.Survey research among substantial numbersof residents enabled us to study the socialcapital levels of the stayers, the moverswithin the neighbourhood, the movers fromsurrounding neighbourhoods and, finally, thenewcomers.

Our first research question dealt with theaverage social capital levels in each group.Contrary to our expectations, we find thatthe newcomers enjoy (access to) relativelyhigh levels of social capital, compared withstayers and movers. While stayers scoredhighest in Hoogvliet Northwest, stayers inDe Horsten have much less access to socialcapital than the movers and the newcomers.In both areas, movers from surrounding neigh-bourhoods are just behind the newcomers intheir social capital scores. Altogether, theseresults imply that length of residence is not adecisive determinant of social capital.

Our second research question asked to whatextent socioeconomic characteristics, neigh-bourhood perceptions and housing aspectsare related to residents’ social capital. Itturns out that households with children andmiddle or higher net incomes in single-family dwellings score relatively high onsocial capital. They are often sociallyupwardly mobile households that made a posi-tive choice to live in the research areas (seePriemus, 2004). Exactly this type of house-hold is represented most among the newco-mers and the least among the stayers. Of allthe socioeconomic characteristics, income inparticular is a strong predictor of socialcapital. This is probably due to a strong con-nection with cultural capital (education).Income also partly determines the ability toaccess support and other resources fromsources other than the neighbourhood. More-over, living in a single-family dwelling andbeing a homeowner is associated with(access to) higher social capital levels; also,the higher residents’ place attachment andperceived neighbourhood quality, the higherthe social capital.

Apart from the factors mentioned above,the newcomers’ relatively high social capitallevels are also related to other factors. Manynewcomers and also movers from the sur-rounding neighbourhoods arrived at the sametime in the neighbourhood, especially in newdwellings. They make a joint new start.Additionally, newcomers are the least hetero-geneous group of all resident categories whenit comes to socioeconomic and householdcharacteristics. Having many people withsimilar characteristics living nearby makespositive social interactions easier and morelikely (see Crawford, 2006, p. 963; Dekkerand Bolt, 2005, p. 2461). Together, thesefactors encourage public familiarity andmutual understanding—however fleeting andsuperficial—and give newcomers, especiallythose with middle and higher householdincomes, a social capital ‘head start’ overthe low-income-groups. Simultaneously, thelack of socioeconomic homogeneity com-bined with continually high levels of occu-pancy turnover explain the lower social




capital scores of stayers in De Horsten, com-pared with stayers in Hoogvliet Northwest.

Our third research question dealt with thepossible connection between social capitaland residents’ expected future length of resi-dence. We could not find a significantrelationship. In other words, residents intend-ing to move within a few years do not havelower levels of social capital than residentswithout moving plans. This adds up to ourearlier finding that length of residence (uptill now) is not a decisive determinant ofsocial capital. Thus, we conclude that the sig-nificance of length of residence in theories onsocial capital, social disorganisation and col-lective efficacy (see Shaw and MacKay,1942; Sampson et al., 1997) needs to berethought, at least for contexts which haveundergone dramatic changes in the past 10years. Length of residence may become animportant factor in the formation of socialcapital above a certain threshold. Such athreshold was clearly not crossed in ourstudy areas, but further research may explorethis possibility.

An obvious shortcoming of this study is itscross-sectional nature. We cannot trace thetrue nature of the links between social capital,place attachment and perceived neighbourhoodquality, or the development of the differencesbetween the resident categories. Nevertheless,the empirical evidence clearly points at signifi-cant associations between several socioeco-nomic, housing and perception indicators, andsocial capital. However, we still know little ofthe interplay between social, economic andcultural capital. We do not know which formof capital is a driver behind processes ofinequality, power differences and social dis-crimination (see Piachaud, 2002; Silva andEdwards, 2004). In sum, theories of socialcapital in relation to neighbourhoods need tobe further developed and tested on the interplaybetween different capital forms.

Even though we did not study the effects ofurban restructuring per se, the findings seemto point to several policy implications. First,the research appears to confirm that providingattractive housing career opportunities formovers within the neighbourhood is a sensible

strategy from a social capital viewpoint (seeDekker and Bolt, 2005, p. 2467). The socialcapital levels of movers within the neighbour-hood are higher than or comparable with thoseof long-term stayers. This seems to suggestthat their access to social capital is not dis-turbed by their intraneighbourhood move(see Piachaud, 2002, pp. 17–18). Lackinglongitudinal data, unfortunately, we cannotproperly test this hypothesis. Secondly, itappears that demolition and new constructionmay improve preconditions for the (re)pro-duction of social capital (see Flint andKearns, 2006, p. 52). Urban restructuringusually results in the physical upgrading ofthe neighbourhood and the provision of attrac-tive housing career opportunities—i.e. new,single-family, owner-occupied dwellings thatmay attract middle-income families fromoutside and within the same neighbourhood(see for example, van Beckhoven and vanKempen, 2003; Kleinhans, 2005; Ministerievan VROM, 2000). Moreover, local auth-orities and housing associations can do muchin terms of social management, such asdealing with nuisance caused by problematictenants, mediation between quarrelling neigh-bours and supporting resident associations.All this can win back or raise the trust of resi-dents in institutions governing the neighbour-hood (Burns et al., 2001; Crawford, 2006;Lelieveldt, 2004). Yet, this is no guaranteefor success. Residents themselves must makeefforts to create social capital. They caninvest in social capital through cursory, every-day social interactions that enable public fam-iliarity and basic levels of trust, which supporta favourable social climate in restructuredneighbourhoods. Finally, restructuring andneighbourhood maintenance policies mustalso ensure attention to parts of neighbour-hoods that were not subject to demolition,new construction and upgrading. Strong andvisible inequalities within neighbourhoodsmay not only result in social disorganisationbut may also stimulate feelings of relativedeprivation among the less affluent residentsin old dwellings (see Taylor and Covington,1988). The case of De Horsten shows thedanger of stayers becoming a neglected




group with high levels of dissatisfaction andlow levels of social capital. They are leastlikely to profit from restructuring, in termsof their housing situation. Obvious physicalcleavages in restructured areas are a breed-ing-ground for social cleavages that hamperthe potential for positive cursory interactionsand public familiarity between residents onthe boundary lines between different blocks.

Notes

1. We are indebted to Talja Blokland for import-ant suggestions on this issue.

2. Strictly speaking, ethnic minority is not aproper term, as it indicates a disadvantagedposition. However, ‘foreign-born’ wouldunjustly exclude the ‘second-generation’immigrants, who are born in The Netherlands,but have at least one parent born abroad.

3. Student’s t ¼ 1.37; df ¼ 869, p ¼ 0.17.4. We also analysed a model without the inter-

action effects. This yielded the same resultsas in Table 4, with no differences in the vari-ables that turn out significant (see Kleinhans,2005, p. 182). However, we here prefer themodel including interaction effects, as theyunderline the bivariate results from Table 3.

5. The explained variance R2 increases from 0.17in the second model to 0.39 in the third model.

6. The number of households with a net house-hold income considered as high (E 3000 permonth or more) amounts to only 12 per centin De Horsten and 7 per cent in Hoogvliet-Northwest.

References

ALLEN, T. (2000) Housing renewal—doesn’t itmake you sick?, Housing Studies, 15(3), pp.443–461.

BECKHOVEN, E. VAN and KEMPEN, R. VAN (2003)Social effects of urban restructuring: a casestudy in Amsterdam and Utrecht, The Nether-lands, Housing Studies, 18(6), pp. 853–875.

BELLAIR, P. E. (1997) Social interaction andcommunity crime: examining the importanceof neighbor networks, Criminology, 35, pp.677–703.

BLOKLAND, T (2003) Urban Bonds. SocialRelationships in an Inner City Neighbourhood.Cambridge: Polity Press.

BOURDIEU, P. (1986) The forms of capital, in:J. G. RICHARDSON (Ed.) Handbook of Theoryand Research for Sociology of Education,pp. 241–258. New York: Greenwood Press.

BOXTEL, R. VAN (2000) Voortgang Grotestedenbe-leid [Progress in the major cities policy]. Letterto the Second Chamber of Parliament,GSB2000/U60445. Ministry of the Interior andKingdom Relations, The Hague.

BREHM, J. and RAHN, W. (1997) Individual-levelevidence for the causes and consequences ofsocial capital, American Journal of PoliticalScience, 41(3), pp. 999–1023.

BRIDGE, G. (2002) The neighbourhood and socialnetworks. Paper No. 4, Centre for Neighbour-hood Research (http://www.neighbourhoodcentre.org.uk).

BRIGGS, X. S. (1998) Brown kids in white suburbs:housing mobility and the many faces of socialcapital,HousingPolicyDebate, 9(1), pp. 177–221.

BROWN, B. B. and PERKINS, D. D. (1992) Disrup-tions in place attachment, in: I. ALTMAN, andS. LOW (Eds) Place Attachment, pp. 279–304.New York: Plenum.

BROWN, B., PERKINS, D. D. and BROWN, G. (2003)Place attachment in a revitalizing neighborhood:individual and block levels of analysis, Journal ofEnvironmental Psychology, 23, pp. 259–271.

BURNS, D., FLINT, J., KEARNS, A. and FORREST, R.(2001) Developing social capital and empower-ing communities: a study of housing associationimpacts. Paper presented at the ENHR Confer-ence, Pultusk, Poland, June.

BUTLER, T. and ROBSON, G. (2001) Social capital,gentrification and neighbourhood change inLondon: a comparison of three south Londonneighbourhoods, Urban Studies, 38(12), pp.2145–2162.

CAMPBELL, K. E. and LEE, B. A. (1992) Sources ofpersonal neighborhood networks: social inte-gration, need or time?, Social Forces, 70(4),pp. 1077–1100.

CLAMPET-LUNDQUIST, S. (2004) HOPE VI reloca-tion: moving to new neighbourhoods and build-ing new ties, Housing Policy Debate, 15(2),pp. 415–448.

COLEMAN, J. S. (1988) Social capital and the cre-ation of human capital, American Journal ofSociology, 94(Supplement), pp. S95–S120.

COLEMAN, J. S. (1990) Foundations of SocialTheory. Cambridge, MA: Belknap Press.

CRAWFORD, A. (2006) ‘Fixing broken promises?’Neighbourhood wardens and social capital,Urban Studies, 43(5/6), pp. 957–976.

DANTAS, A. M. (1988) Overspill as an alternativestyle of gentrification: the case of Riverdale,Toronto, in: T. E. BUNTING and P. FILION

(Eds) The changing Canadian inner city,pp. 73–86. Publication Series 31, Departmentof Geography, University of Waterloo, Ontario.

DAVIDSON, W. R. and COTTER, P. R. (1986)Measurement and sense of community withinthe sphere of the city, Journal of AppliedSocial Psychology, 16, pp. 608–619.




DEKKER, K. and BOLT, G. (2005) Social cohesionin post-war estates in The Netherlands:differences between socioeconomic and ethnicgroups, Urban Studies, 42(13), pp. 2447–2470.

DEKKER, S. (2005) Voortgang actieprogrammaherstructurering [Progress in the actionprogramme urban restructuring] Letter to theSecond Chamber of Parliament, DGW/SR2005019562. Ministry of Housing, SpatialPlanning and the Environment, The Hague.

DIPASQUALE, D. and GLAESER, E. (1999) Incen-tives and social capital: are homeowners bettercitizens?, Journal of Urban Economics, 45(2),pp. 354–384.

DRUKKER, M., FERON, F. and OS, J. VAN (2005) Debuurt doet ertoe: Maastricht kwaliteit van levenstudie [The neighbourhood matters: Maastrichtquality of life study]. Universiteit Maastricht.

DUNCAN, T. E., DUNCAN, S. C., OKUT, H. ET AL.

(2003) A multilevel contextual model of neigh-bourhood collective efficacy, American Journalof Community Psychology, 32(3/4), pp.245–252.

EKSTROM, M. (1994) Elderly people’s experiencesof housing renewal and forced relocation: socialtheories and contextual analysis in explanationsof emotional experiences, Housing Studies, 9,pp. 369–391.

ELIAS, N. and SCOTSON, J. L. (1965), The Estab-lished and the Outsiders: A SociologicalEnquiry into Community Problems. London:Frank Cass & Co. Ltd.

ELLAWAY, A. S., MACINTYRE, S. and KEARNS, A.(2001) Perceptions of place and health insocially contrasting neighbourhoods, UrbanStudies, 38(12), pp. 2299–2316.

ELSINGA, M. and HOEKSTRA, J. (2004) De beteke-nis van eigenwoningbezit [The meaning ofhomeownership]. VROM-Raad, The Hague.

FIELD, J. (2003) Social Capital. London: Routledge.FINE, B. (2001) Social Capital versus Social

Theory: Political Economy and Social Scienceat the Turn of the Millennium. London:Routledge.

FISCHER, C. S. (1982) To Dwell Among Friends:Personal Networks in Town and City. Chicago,IL: University of Chicago Press.

FLINT, J. and KEARNS, A. (2006) Housing, neigh-bourhood renewal and social capital: the caseof registered social landlords in Scotland, Euro-pean Journal of Housing Policy, 6(1), pp.31–54.

FOLEY, M. and EDWARDS, B. (1999) Is it time todisinvest in social capital?, Journal of PublicPolicy, 19, pp. 141–173.

FORREST, R. and KEARNS, A. (1999) Joined-upPlaces? Social Cohesion and NeighbourhoodRegeneration. York: Joseph RowntreeFoundation.

FORREST, R. and KEARNS, A. (2001) Social cohe-sion, social capital and the neighbourhood,Urban Studies, 38(12), pp. 2125–2143.

FRIED, M. (1967) Functions of the working-classcommunity in modern urban society: impli-cations for forced relocation, Journal of theAmerican Institute of Planners, 33, pp. 90–103.

FUKUYAMA, F. (1995) Trust: The Social Virtuesand the Creation of Prosperity. New York:The Free Press.

FUKUYAMA, F. (1999) The Great Disruption:Human Nature and the Reconstitution of SocialOrder. London: Profile Books.

GANS, H. J. (1991) People, Plans, and Policies:Essays on Poverty, Racism, and Other NationalUrban Problems. New York: Columbia Univer-sity Press.

GITTELL, R. and VIDAL, A. (1998) CommunityOrganizing: Building Social Capital as a Devel-opment Strategy. Thousand Oaks, CA: Sage.

GOETZ, E. G. (2002) Forced relocation vs.voluntary mobility: the effects of dispersalprogrammes on households, Housing Studies,17, pp. 107–123.

GRANOVETTER, M. S. (1973) The strength of weakties, American Journal of Sociology, 78(6),pp. 1360–1380.

GROOTAERT, C., NARAYAN, D., NYHAN-JONES, V.and WOOLCOCK, M. (2002) Integrated question-naire for the measurement of social capital.Social Capital Thematic Group, The WorldBank (http://poverty.worldbank.org/library).

HALPERN, D. (2005) Social Capital. Cambridge:Polity Press.

HENNING, C. and LIEBERG, M. (1996) Strong tiesor weak ties? Neighbourhood networks in anew perspective, Scandinavian Housing andPlanning Research, 13, pp. 3–26.

JUPP, B. (1999) Living Together: Community Lifeon Mixed Tenure Estates. London: Demos.

KASARDA, J. D. and JANOWITZ, M. (1974) Com-munity attachment in mass society, AmericanSociological Review, 39, pp. 328–339.

KEARNS, A. (2004) Social capital, regenerationand urban policy. Paper No. 25, Centre forNeighbourhood Research (http://www.neighbourhoodcentre.org.uk).

KEMPEN, R. VAN and PRIEMUS, H. (1999) Undi-vided cities in the Netherlands: present situationand political rhetoric, Housing Studies, 14(5),pp. 641–657.

KEMPEN, R. VAN and PRIEMUS, H. (2002) Revolu-tion in social housing in the Netherlands: poss-ible effects of new housing policies, UrbanStudies, 39(2), pp. 237–253.

KLEINHANS, R. (2003) Displaced but still movingupwards in the housing career? Implications offorced residential relocation in The Netherlands,Housing Studies, 18(4), pp. 473–499.




KLEINHANS, R. (2004) Social implications ofhousing diversification in urban renewal: areview of recent literature, Journal of Housingand the Built Environment, 19(4), pp. 367–390.

KLEINHANS, R. (2005) Sociale Implicaties vanHerstructurering en Herhuisvesting [Socialimplications of urban restructuring and forcedrelocation]. Delft: Delft University Press.

LELIEVELDT, H. (2004) Helping citizens helpthemselves: neighborhood improvement pro-grams and the impact of social networks, trustand norms on neighborhood-oriented forms ofparticipation, Urban Affairs Review, 39(5), pp.531–551.

LIN, N. (2001) Social Capital: A Theory of SocialStructure and Action. Cambridge: CambridgeUniversity Press.

LOFLAND, L. (1985) A World of Strangers.Prospect Heights, IL: Waveland Press Inc.

MIDDLETON, A., MURIE, A. and GROVES, R.(2005) Social capital and neighbourhoods thatwork, Urban Studies, 42(10), pp. 1711–1738.

MINISTERIE VAN VROM (1997) Nota stedelijkevernieuwing [Policy memorandum urbanrenewal]. Ministry of Housing, Spatial Planningand the Environment, The Hague.

MINISTERIE VAN VROM (2000) Mensen wensenwonen: wonen in de 21ste eeuw [What peoplewant, where people live]. Ministry of Housing,Spatial Planning and the Environment, TheHague.

MULDER, C. H. and HOOIMEIJER, P. (1999) Resi-dential relocations in the life course, in: L.VAN WISSEN and P. DYKSTRA (Eds) PopulationIssues: An Interdisciplinary Focus, pp.159–186. New York: Kluwer/Plenum.

NARAYAN, D. and CASSIDY, M. F. (2001) Adimensional approach to measuring socialcapital: development and validation of a socialcapital inventory, Current Sociology, 49(2),pp. 59–102.

OSTENDORF, W., MUSTERD, S. and VOS, S. DE

(2001) Social mix and the neighbourhoodeffect: policy ambitions and empirical evidence,Housing Studies, 16(3), pp. 371–380.

PARKES, A., KEARNS, A. and ATKINSON, R. (2002)Determinants of neighbourhood dissatisfaction.Paper No. 1, Centre for NeighbourhoodResearch (http://www.neighbourhoodcentre.org.uk).

PERKINS, D. D. and LONG, D. A. (2002) Neighbor-hood sense of community and social capital: amulti-level analysis, in: A. FISHER, C. SONN

and B. BISHOP (Eds) Psychological Sense ofCommunity: Research and Implications, pp.291–318. New York: Kluwer AcademicPublishers.

PIACHAUD, D. (2002) Capital and the determinantsof poverty and social exclusion. Paper No. 60,

Centre for Analysis of Social Exclusion,London School of Economics.

POPP, H. (1976) The residential location process.some theoretical and empirical considerations,Tidschrift voor Economische en Sociale Geogra-fie, 67(5), pp. 300–305.

PORTES, A. (1998) Social capital: its origins andapplications in modern sociology, AnnualReview of Sociology, 24, pp. 1–24.

PORTES, A. and LANDOLT, P. (1996) The downsideof social capital, American Prospect, 26,pp. 18–21.

PRIEMUS, H. (2004) The path to successful urbanrenewal: current policy debates in The Nether-lands, Journal of Housing and the Built Environ-ment, 19(3), pp. 199–209.

PUTNAM, R. D. (1993) Making Democracy Work:Civic Tradition in Modern Italy. Princeton, NJ:Princeton University Press.

PUTNAM, R. D. (2000) Bowling Alone: The Col-lapse and Revival of American Community.New York: Simon & Schuster.

REIJNDORP, A., KOMPIER, V., METAAL, S. ET AL.

(1998) Buitenwijk. Stedelijkheid op afstand[Suburb: urbanity at a distance]. Rotterdam:NAi Uitgevers.

ROSS, C. E., MIROWSKY, J. and PRIBESH, S. (2001)Powerlessness and the amplification of threat:neighbourhood disadvantage, disorder and mis-trust, American Sociological Review, 66,pp. 568–591.

SAEGERT, S. and WINKEL, G. (2004) Crime, socialcapital, and community participation, AmericanJournal of Community Psychology, 34(3/4),pp. 219–233.

SAMPSON, R. J., RAUDENBUSH, S. W. and EARLS,F. (1997) Neighbourhoods and violent crime: amultilevel study of collective efficacy, Science,277, pp. 918–924.

SHAW, C. R. and MACKAY, H. D. (1942) JuvenileDelinquency in Urban Areas. Chicago, IL: Uni-versity of Chicago Press.

SILVA, E. and EDWARDS, R. (2004) Operationaliz-ing Bourdieu on capitals: a discussion on theconstruction of the object. Working PaperNo. 7, ESRC Research Methods Programme,Open University/South Bank University, London.

SKJAEVELAND, O. and GARLNG, T. (1996) A multi-dimensional measure of neighboring, AmericanJournal of Community Psychology, 24(3),pp. 413–435.

STEVENS, J. (1996) Applied Multivariate Statisticsfor the Social Sciences. Mahwah, NJ: LawrenceErlbaum Associates.

TAYLOR, R. and COVINGTON, J. (1988) Neighbor-hood changes in ecology and violence, Crimi-nology, 26, pp. 553–589.

TEMKIN, K. and ROHE, W. M. (1998) Social capitaland neighbourhood stability: an empirical




investigation, Housing Policy Debate, 9(1),pp. 66–68.

TUNSTALL, R. (2003) ‘Mixed tenure’ policy in theUK: privatisation, pluralism or euphemism?,Housing, Theory and Society, 20(3),pp. 153–159.

UITERMARK, J. (2003) ‘Social mixing’ and themanagement of disadvantaged neighbourhoods:the Dutch policy of urban restructuring revisited,Urban Studies, 40(3), pp. 531–549.

VALE, L. (1997) Empathological places: residents’ambivalence towards remaining in publichousing, Journal of Planning and EducationResearch, 16, pp. 159–175.

WELLMAN, B., CARRINGTON, P. and HALL, A.(1988) Networks as personal communities, in:B. WELLMAN and S. BERKOWITZ (Eds) SocialStructures: A Network Approach, pp. 130–184.Cambridge: Cambridge University Press.

WOOD, M. (2003) A balancing act? Tenure diversi-fication in Australia and the UK, Urban Policyand Research, 21(1), pp. 45–56.

WOOLCOCK, M. (1998) Social capital and econ-omic development: towards a theoreticalsynthesis and policy framework, Theory andSociety, 27, pp. 151–208.

WOUDEN, R. VAN DER and BRUIJNE, E. DE (2001)The city and the urban region: problems andperspectives of the four major urban areas inthe Randstad. Sociaal en Cultureel Planbureau,The Hague.

WRR (WETENSCHAPPELIJKE RAAD VOOR HET

REGERINGSBELEID) [Scientific Council forGovernment Policy] (2005) Vertrouwen in deBuurt [Trust in the neighbourhood]. Amsterdam:Amsterdam University Press.

Appendix. The Social Capital Index

Below is a list of all social capital indicators, asused in our survey. Many indicators are derivedfrom validated social capital surveys (such asGrootaert et al., 2002).

(1) In this neighbourhood, we are on good termswith each other.

(2) I must solve many problems for myselfbecause few people support me.

(3) If I help a neighbour with something, Iexpect him to return a favour in the future.

(4) It is not easy to establish contacts with thepeople around here.

(5) In case of emergency, I can always asksomeone in this neighbourhood for help.

(6) There are tensions here between newcomersand people who have lived here for a longtime.

(7) Actual support offered to neighbours duringthe past two months.

(8) Active membership in a voluntary associ-ation (resident organisation, sport club,church, and other).

(9) Voluntary work in an association or ingeneral.

(10) Co-operation with other residents in the pastyear to achieve something for theneighbourhood.

(11) The people around here would co-operatewell to get something done for the neigh-bourhood, such as a face-lift of the publicpark.

(12) In this neighbourhood, there is a good levelof social control.

(13) The residents in this neighbourhood take noaccount of each other.

(14) I feel jointly responsible for the liveability inthis neighbourhood.

(15) The residents have common norms withregard to keeping this neighbourhood tidy.

(16) Residents should not meddle with eachother’s affairs.

(17) If you encounter a person in this area, wouldyou know if he or she lives in thisneighbourhood?

(18) If a resident parks his car on the sidewalk,would you ask him to move it to a parkingplace?

(19) Generally speaking, residents in this neigh-bourhood can be trusted.

(20) When I go on a holiday, I can leave my housekey safely with my neighbours or otherresidents.

(21) One cannot be too careful in dealing withpeople you do not know.

(22) I don’t mind several ethnic groupsliving in this neighbourhood alongside eachother.

Items (2), (3), (4), (6), (13) and (21) have areversed meaning and are recoded accordinglyItems (7), (8), (9) and (10) are bivariate items(0 ¼ no; 1 ¼ yes).

A principal components analysis has indicatedthree relevant components of social capital, eachwith an eigen value of more than 1 (Kaisercriterion; see Stevens, 1996, p. 367)

—C1 social interactions: variables (1), (2), (4), (5),(7), (13) and (20) (Cronbach’s a ¼ 0.73).

—C2 norms and trust: variables (3), (11), (12),(15), (17) and (19) (Cronbach’s a ¼ 0.61).




—C3 associational activity: variables (8), (9) and(10) (Cronbach’s a ¼ 0.56).

The remaining six variables are joined in threepairs. However, a factor or component with only

two variables is, strictly speaking, not a factor(Stevens, 1996, p. 373). Consequently, these ‘com-ponents’ are not analysed separately, but thematching variables are adequately included in theoverall social capital index.




understanding social capital in recently restructured urban neighbourhoods: two case studies in...

Documents