Intergenerational transmission of neighbourhood poverty: an analysis of neighbourhood histories of...

Intergenerational transmission of neighbourhoodpoverty: an analysis of neighbourhood historiesof individuals

Maarten van Ham1,2,3, Lina Hedman4, David Manley5,Rory Coulter6 and John Östh7,8

The extent to which socioeconomic (dis)advantage is transmitted between generations is receiving increasingattention from academics and policymakers. However, few studies have investigated whether there is a spatialdimension to this intergenerational transmission of (dis)advantage. Drawing on the concept of neighbourhoodbiographies, this study contends that there are links between the places individuals live with their parents and theirsubsequent neighbourhood experiences as independent adults. Using individual-level register data tracking thewhole Stockholm population from 1990 to 2008, and bespoke neighbourhoods, this study is the first to usesequencing techniques to construct individual neighbourhood histories. Through visualisation methods andordered logit models, we demonstrate that the socioeconomic composition of the neighbourhood children lived inbefore they left the parental home is strongly related to the status of the neighbourhood they live in 5, 12 and 18years later. Children living with their parents in high poverty concentration neighbourhoods are very likely to endup in similar neighbourhoods much later in life. The parental neighbourhood is also important in predicting thecumulative exposure to poverty concentration neighbourhoods over a long period of early adulthood. Ethnicminorities were found to have the longest cumulative exposure to poverty concentration neighbourhoods. Thesefindings imply that for some groups, disadvantage is both inherited and highly persistent.

Key words intergenerational transmission; deprived neighbourhoods; neighbourhood biography; sequence anal-ysis; Sweden

1OTB – Research for the Built Environment, Faculty of Architecture and the Built Environment, Delft University of Technology, Delft, TheNetherlandsEmail: [email protected] of St Andrews, UK3IZA Bonn, Germany4Institute for Housing and Urban Research, Uppsala University, Uppsala, Sweden5School of Geographical Sciences, University of Bristol, Bristol, BS8 1SS6Department of Sociology, University of Cambridge, Cambridge, CB2 3RQ7Department of Social and Economic Geography, Uppsala University, Uppsala, Sweden8Department of Human Geography, Stockholm University, Stockholm, Sweden

Revised manuscript received 17 June 2013

Introduction

There is a large body of literature investigating theintergenerational transmission of (dis)advantage. It haslong been recognised that ‘the fortunes of children arelinked to their parents’ (Becker and Tomes 1979, 1153),and it is now well established that individual charac-teristics, such as labour market earnings and educa-tional attainment, correlate strongly between parentsand children (D’Addio 2007). The extent to whichsocioeconomic (dis)advantage is transmitted betweengenerations is receiving increasing attention frompolicymakers. According to the UK government reportOpening doors, breaking barriers: a strategy for socialmobility,

In Britain today, life chances are narrowed for too many bythe circumstances of their birth: the home they’re born into,the neighbourhood they grow up in or the jobs their parentsdo. Patterns of inequality are imprinted from one generationto the next. (Nick Clegg, Cabinet Office 2011, 3)

This liberal belief in the importance of severing thelinks between ascribed or inherited characteristics andindividual attainment has become an important policyobjective across the developed world, advocated forboth equity and efficiency reasons (OECD 2010).

It has been suggested – sometimes implicitly – thatthe intergenerational transmission of (dis)advantagealso has a spatial dimension (Duncan and Raudenbush2001; Jencks and Mayer 1990; Sampson and Wilson

The information, practices and views in this article are those of the author(s) and do not necessarily reflect the opinion ofthe Royal Geographical Society (with IBG). ISSN 0020-2754 Citation: 2014 39 402–417 doi: 10.1111/tran.12040

© 2013 The Authors. Transactions of the Institute of British Geographers published by John Wiley & Sons Ltd on behalf ofRoyal Geographical Society (with The Institute of British Geographers)

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use anddistribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

Copyright line has been changed since first publication on 29 October 2013.

1995; van Ham et al. 2012). The idea is that childrenwho grow up in poverty concentration neighbourhoodsmight be more likely than others to end up in such aneighbourhood as adults. To our knowledge there isonly one study to date that empirically investigates thisspatial dimension (see Vartanian et al. 2007).

The possibility of a spatial dimension to intergener-ational transmission of (dis)advantage is highly relevantfor the literature on neighbourhood effects. Thisliterature investigates whether living in a povertyconcentration neighbourhood (see Wilson 1987) has anegative effect on residents’ life chances (related to, forexample, income, education and health), over andabove the effect of their individual characteristics (seeEllen and Turner 1997; Galster 2002; Manley and vanHam 2012). Despite an enormous and growing body ofliterature on neighbourhood effects, there is littleagreement on the causal mechanisms that mightproduce them, their relative importance comparedwith individual characteristics, and under which cir-cumstances and where these effects are important (vanHam et al. 2012). A major problem in identifying causalneighbourhood effects is that people do not randomlyselect their neighbourhoods, and as a result parameterestimates for these effects are biased (Durlauf 2004;van Ham and Manley 2010). This process of neigh-bourhood selection, which over the life course cumu-latively creates an individual’s neighbourhoodbiography, may be influenced by the parental neigh-bourhood. Consequently, individual life outcomesmight not only be related to the current neighbour-hood, but also to neighbourhood histories.

Insight into the neighbourhood histories of individ-uals will not only benefit the literature on neighbour-hood effects, but will also contribute to ourunderstanding of both segregation and residentialmobility processes. Many studies of residential mobilityuse the life course approach as a starting point andanalyse longitudinal data, but few studies investigatetrue life courses empirically as they only investigatetransitions in states between two years (e.g. Rabe andTaylor 2010). As a result, very little is known about thewider neighbourhood biographies within which theseevents and transitions are situated. This is problematic,as the biographical context within which an eventoccurs can condition its significance and meaning(Coulter and van Ham forthcoming; Dykstra and vanWissen 1999). For example, a move from an affluentneighbourhood to a poverty concentration neighbour-hood has a very different meaning depending onwhether the stay in this poor neighbourhood is tempo-rary or more permanent.

This is the first paper to construct the entireneighbourhood histories of a large group of individualsover a long period of time. We investigated theintergenerational transmission of neighbourhood pov-

erty in Stockholm through the effect of the parentalneighbourhood on individual neighbourhood biogra-phies over a period of almost two decades. This studyhypothesised that the parental neighbourhood haspredictive value for neighbourhood outcomes later inlife and for the cumulative exposure to povertyconcentration neighbourhoods over the life course. Torelate the neighbourhood careers of parental homeleavers in the Stockholm metropolitan area to the lastneighbourhood they lived in with their parents, we used1990–2008 longitudinal register data from the GeoS-weden database. We defined poverty concentrationneighbourhoods based on the percentage of low-income neighbours in the local area. We used bespokeneighbourhoods based on the characteristics of the 500people living closest to each individual (Östh et al.forthcoming) and used innovative sequencing tech-niques to visualise individual neighbourhood histories(for a residential mobility application see Coulter andvan Ham forthcoming). Sweden provides an excellentcase study for the analysis of the intergenerationaltransmission of neighbourhood characteristics, becausedespite evidence for increasing social polarisation(Hedin et al. 2012), Sweden is widely considered tobe one of the least stratified Western societies (on thenature of welfare states in Western societies seeEsping-Andersen 1990). As a result, this study providesan important counterweight to studies on US andBritish societies that dominate the segregation litera-ture (Maloutas 2012).

Literature review

To link our understanding of residential mobility(histories) and neighbourhood choice to the literatureon intergenerational transmission of (dis)advantage,this study adopts the life course approach as used inmany residential mobility studies (Clark and Huang2003). In contrast with the more normative anddeterministic life cycle approach, life course theoryargues that individuals experience their own uniquesequence of life events as they age (Clark andDieleman 1996; Geist and McManus 2008; van Ham2012). As a result, individual lives can be thought of asunique personal biographies (Dykstra and van Wissen1999; Elder 1994). Mulder and Hooimeijer (1999)argue that the life events occurring within thesepersonal biographies can be grouped into four parallellife course careers. In this framework, it is the sequenceof events experienced in these interlinking household,labour force, education and housing careers thatinfluence an individual’s residential mobility behaviourthroughout their lifetime (Clark and Withers 2007).Crucially, the life course model posits that an individ-ual’s choices and behaviours can be strongly affected bythe events or states they have experienced earlier in

Intergenerational neighbourhood poverty 403

ISSN 0020-2754 Citation: 2014 39 402–417 doi: 10.1111/tran.12040© 2013 The Authors. Transactions of the Institute of British Geographers published by John Wiley & Sons Ltd on behalf of

Royal Geographical Society (with The Institute of British Geographers)

their life (Dykstra and van Wissen 1999; Feijten 2005;Feijten et al. 2008). Adopting a life course approachtherefore guides researchers to analyse the occurrenceof events within the long-term individual biography andmacro-context within which these are situated (Aisen-brey and Fasang 2010).

While the evolution of housing careers across thelife course has been a focus for much housing andmobility research (Clark et al. 2003; Feijten and Mul-der 2005), most studies have been based around theempirical analysis of either cross-sectional data con-taining some retrospective information, or short peri-ods of longitudinal data (e.g. Clark and Ledwith 2006;Geist and McManus 2008). The result is a focus onyear-to-year mobility instead of mobility biographiesover the life course. Such studies typically show thathouseholds often move between dwellings to adjusttheir space consumption in response to their changinghousehold needs (Clark and Dieleman 1996; Clark andHuang 2003). Changes in household composition arealso often associated with moves to different types ofdwellings in different types of neighbourhoods (Mulderand Hooimeijer 1999). In general, younger adults andsingles prefer to live in neighbourhoods located closerto jobs and amenities in city centres, while householdswith children prefer to live in suburban locations (seeKim et al. 2005). Only a few studies of residentialmobility and housing careers have focused on con-structing and analysing individual mobility biographiesover longer periods of time, using long-running panelsurveys (e.g. Coulter and van Ham forthcoming;Pollock 2007; Stovel and Bolan 2004). These studieshave outlined new ways of conceptualising and con-structing housing biographies, either as visual timelines(Coulter and van Ham forthcoming) or by usingoptimal matching methods to identify clusters of similarresidential histories (Pollock 2007; Stovel and Bolan2004).

Given that residential mobility involves neighbour-hood as well as dwelling selection, it is surprising that fewstudies have extended the life course approach toinvestigate neighbourhood biographies. An increase insocio-economic status over the life course has beenfound to promote mobility into a more affluent neigh-bourhood (Rossi 1980; South andCrowder 1997; see alsoClark et al. 2003). Vartanian et al. (2007) noted that atthe same time welfare receipt, public housing (Kasarda1988) and homeownership (South and Crowder 1997)have been found to be limiting mobility into betterneighbourhoods. In general it can therefore be expectedthat an analysis of neighbourhood biographies will showupward trajectories of neighbourhood status acrossindividual lives. However, there are indications that thismight not be equally true for all ethnic groups (Simpsonand Finney 2009). Generally speaking, ethnic minoritieslive in significantly worse neighbourhoods than ethnic

majority groups (Crowder and South 2005; Massey et al.1994; Quillian 2003) and ethnic minorities are less likelyto translate human capital into upward residentialmobility (South and Deane 1993; Vartanian et al. 2007).

The few studies that have analysed the types ofplaces in which individuals live across their life coursehave tended to distinguish places according to theirpopulation size or physical characteristics rather thantheir socioeconomic composition. For instance, Stoveland Bolan (2004) distinguish nine ‘place-types’ rangingfrom sparsely populated rural areas to large metropol-itan centres. In contrast, Feijten et al. (2008) classifiedneighbourhoods according to whether they werelocated in central city, suburban or rural areas. Thisstudy found that the neighbourhood in which anindividual grew up was related to the types of placesthey lived in later in life (see also Blaauboer 2011).Feijten et al. argue that this may be because childrenare socialised into preferring a similar type of neigh-bourhood to that in which they grew up. Similararguments have been put forward to explain theintergenerational transmission of dwelling preferences(see Helderman and Mulder 2007; Kurz 2004; Mulderand Smits 1998). Hence, children growing up in anowned property are more likely to exhibit preferencesfor homeownership as adults (Henretta 1984).

In a spatially segmented housing market, suchpreferences are also likely to affect neighbourhoodchoice. As stated in the introduction, to our knowledgethere is only one study that explicitly investigated theintergenerational transmission of neighbourhood type(Vartanian et al. 2007). This study used sibling datafrom the Panel Study of Income Dynamics linked withUS Census data. Their results confirmed the hypothesisthat childhood neighbourhood disadvantage has nega-tive effects on adult neighbourhood quality for thoseliving in the lowest quality neighbourhoods (Vartanianet al. 2007). They argue that family poverty and thelikelihood of residing in disadvantaged neighbourhoodsis inherited across generations (cf. Henretta 1984). Thismeans that children who grow up in poor neighbour-hoods are more likely to reside in similar environmentsas adults. Vartanian and colleagues explain theirfindings with neighbourhood effects theory. Theysuggest that children growing up in poverty areas willexperience negative neighbourhood effects on theirincome and employment opportunities, limiting theirsubsequent options in the housing market as anindependent adult. Furthermore, growing up in apoverty neighbourhood may result in negative effectson their perceptions of their future possibilities ofmoving to more advantaged neighbourhoods, whichmay make them even more likely to remain in povertyneighbourhoods as adults.

There are a number of cultural reasons why the typeof neighbourhood an individual lived in with their

404 Maarten van Ham et al.

ISSN 0020-2754 Citation: 2014 39 402–417 doi: 10.1111/tran.12040© 2013 The Authors. Transactions of the Institute of British Geographers published by John Wiley & Sons Ltd on behalf ofRoyal Geographical Society (with The Institute of British Geographers)

parents may condition their subsequent neighbourhoodexperiences. A particularly important mechanism forthe transmission of neighbourhood quality could bethrough the inheritance of social norms. Social normsevolve over the life course and are dependent on thetype, number and nature of contacts made betweenpeople and the environment to which they are exposed.In the parent to child relationship, the greatest impactparents can make on their children’s values will bewhile the child is growing up. Thus, the neighbourhoodin which an individual lives as a child could shape theirfuture neighbourhood career. By extension, we canconsider whether or not individuals growing up inneighbourhoods that occupy lower positions in theneighbourhood hierarchy are likely to live in similarneighbourhoods later in life, or if they are able toexperience neighbourhood hierarchy mobility andmove into neighbourhoods with a higher status.

Parents may also socialise their children withincertain groups with the (tacit) expectation that theywill acquire the cultural traits of these groups.Children will also acquire the norms of their parentsas a result of parental actions mediated through theirimmediate social environment. As they grow upchildren learn through interaction and observation,potentially adopting the traits that they then enact inlater life. This could influence their later behaviour, byaltering the types of people they prefer to associatewith, the places that they visit and work and, impor-tantly for this paper, the type of places (neighbour-hoods) that they wish to live in (see Bisin and Verdier1998). In addition, the intergenerational transmissionof earnings, income and educational achievement(D’Addio 2007; Solon 1999) is also likely to link theneighbourhood biographies of children and parents.Given the divergent neighbourhood experiences ofethnic groups, ethnicity may be an important factormediating the intergenerational transmission of neigh-bourhood (dis)advantage.

Based on the above discussion we expect that theneighbourhood biographies of children leaving theparental home will be related to the parental neigh-bourhood status. More precisely, we expect that chil-dren from poverty concentration neighbourhoods aremore likely to sort into poor neighbourhoods as adultsthan children from more affluent neighbourhoods. Weexpect the intergenerational transmission of neighbour-hood disadvantage to be especially strong for thosewith parents living in the poorest neighbourhoods. Wealso expect the transmission effects to be stronger forethnic minority children than for ethnic majoritychildren. Given that cultural and housing marketfactors may be relevant for transmission processes, weanticipate the effects of parental neighbourhood char-acteristics to persist even after controlling for the lifecourse attainment of children.

Data and methods

The data used for this study are derived from GeoS-weden, a longitudinal micro-database containing theentire Swedish population tracked from 1990 to 2008.The database is constructed from a number of differentannual administrative registers and includes demo-graphic, geographic and socio-economic data for eachindividual living in Sweden. Within this database, it ispossible to follow people over an 18-year period andconstruct their full neighbourhood histories.

In this study we restricted our selection to peopleliving in the Stockholm metropolitan region,1 to ensurethat the definition of ‘neighbourhood’ was as consistentas possible. It is clear that neighbourhoods in the highlyrural far north of Sweden are very different from inner-city neighbourhoods, while two neighbourhoods withinthe Stockholm metropolitan region are more likely tobe a similar size. To some extent, Stockholm can bedescribed as a microcosm representing Swedish society,but on the other hand, Stockholm has some uniquecharacteristics (see Hedin et al. 2012). Stockholm hasrelatively high incomes and housing costs, lower levelsof unemployment and higher levels of job creation thanthe rest of Sweden. The Stockholm metropolitan areaalso stands out with low average ages, higher thanaverage educational levels, greater shares of migrantsborn abroad and low fertility rates. The within Stock-holm variation between neighbourhoods in demo-graphic structure, socio-economic status and housingmarket characteristics is considerable, making Stock-holm a good candidate for studies of intergenerationaltransmission of neighbourhood effects.

To identify home leavers, we restricted the selectionto individuals who were between 16 and 25 years oldand living with their parents in 1990 who had left theparental home by 1991. These selections resulted in atotal of 13 526 parental home leavers for whom we canconstruct neighbourhood histories. It is important tonote that this is the full population of Stockholmmetropolitan region home leavers in 1990–1991, not asample. Age was not included in our multivariatemodels because the effect of age was not significant andbecause including age did not qualitatively affect theparameters of the other variables included.

Instead of using standard administrative neighbour-hoods, we used bespoke neighbourhoods defined usingthe characteristics of the 500 persons living closest toeach individual in the dataset. Using Equipop software(Östh et al. forthcoming) and a dataset with100 9 100-metre geo-coordinates2 (the smallest geo-graphical coding available in the dataset), the charac-teristics of the 500 nearest neighbours (people) werecalculated for each location. We have repeated this foreach year (1990–2008), so neighbourhoods are allowedto change over time.




The advantage of this definition compared withusing standard administrative neighbourhoods is thatthe resulting neighbourhood characteristics are a betterrepresentation of the actual residential environment ofeach individual (avoiding boundary effects). For thisstudy on the intergenerational transmission of neigh-bourhood poverty, the main neighbourhood variable ofinterest was the share of low-income people in thebespoke neighbourhood, where income is defined aspersonal income from work.3 Individuals were categor-ised as having a low income if their income fell into thelowest quintile of the entire Swedish income distribu-tion. Using the Equipop software, we calculated thepercentage of low-income neighbours among the 500nearest neighbours of working age for each residentiallocation. The final step was to create quintiles based onthis neighbourhood characteristic. Neighbourhoods inthe first quintile contain the lowest concentration oflow-income individuals and neighbourhoods in the fifthquintile contain the highest concentration of low-income individuals. Henceforth we refer to thesefifth-quintile neighbourhoods as ‘poverty concentrationneighbourhoods’. Table I gives some basic informationon each of the five neighbourhood quintiles for both1990 and 2008.4 It can be seen that poverty concentra-tion neighbourhoods (quintile 5) have the highestpercentage of low-income people, the highest share ofethnic minorities and the highest share of public rentaldwellings in both years.

The type of neighbourhood where an individual livescan change over time due to residential moves and dueto changes in the neighbourhood composition. Becausethis study focuses on how residential mobility constructsneighbourhood histories, we only allowed the neigh-bourhood quintile of an individual to change after anactual residential move. Given that neighbourhoodsremain relatively constant in the short term (see forinstance Meen et al. 2007), this decision will not biasour results. We measure childhood neighbourhoodexperience by using the last neighbourhood children

lived in before they left the parental home. This islargely a pragmatic decision made to enable us to followpeople for as long as possible. It has been shownpreviously that neighbourhood characteristics are highlycorrelated throughout childhood, so using the lastchildhood neighbourhood should only produce limitedbias (Kuntz et al. 2003; Vartanian et al. 2007).

We then created individual neighbourhood quintilesequences using a visualisation method (for anotherapplication see Coulter and van Ham forthcoming). Toour knowledge this method has not previously beenused to investigate individual neighbourhood histories.Individual sequences were created using SQ-Ados inStata (version 11). These track the ordering of anelement variable over each of the person-years pro-vided by each respondent (Brzinsky-Fay et al. 2006). Inthis case, the element is the neighbourhood incomequintile. The resultant individual neighbourhood his-tories can be combined and visualised as a series ofpersonal timelines (see Figures 1–3). Within theseplots, each horizontal line contains the neighbourhoodhistory of an individual between 1990 and 2008. Thetimeline is colour coded for each of the years based onthe neighbourhood income quintile experienced in thatyear. The first segment in each individual historyrepresents the neighbourhood income quintile of theparental address (remember that the study populationonly includes individuals who left the parental homebetween 1990 and 1991). All subsequent colouredsegments represent the neighbourhoods that peoplelived in after leaving the parental home. If there is nochange of colour from one year to the next, theindividual has not moved, or has relocated but notchanged neighbourhood quality. The GeoSweden dataare particularly suitable for constructing neighbour-hood histories because there is almost no attrition (as itis based on register data), and as a result we were ableto construct neighbourhood histories for the fullpopulation of home leavers in the Stockholm metro-politan region. This would not have been possible using

Table I Descriptives of neighbourhood quintiles (1990 and 2008)

Neighbourhoodquintiles

1990 2008

% low-incomeneighbours

% ethnicminorities % public rentals

% low-incomeneighbours

% ethnicminorities % public rentals

Mean Std Dev. Mean Std Dev. Mean Std Dev. Mean Std Dev. Mean Std Dev. Mean Std Dev.

1 (low poverty) 10.1 1.5 2.2 2.0 6.7 17.0 9.6 1.4 6.4 5.3 4.6 13.72 13.2 0.7 3.8 3.6 20.2 28.3 12.3 0.6 7.2 6.1 8.9 19.43 15.5 0.7 5.2 4.6 34.3 35.6 14.5 0.7 8.3 6.9 11.9 22.84 18.1 0.9 7.2 5.9 48.3 39.9 17.4 1.1 9.7 8.1 14.9 25.95 (povertyconcentration)

24.1 5.4 18.9 15.3 61.5 39.7 24.8 6.2 16.3 15.1 26.4 36.0

Note: The large differences in the share of public rentals in all quintiles between the years 1990 and 2008 are due to tenure transformationswhere public rental dwellings are turned into private rental or cooperative dwellings



panel data, which often exhibit a high rate of partic-ipant attrition in the first years of data collection. Thisattrition results in a high number of incompleteneighbourhood histories (see also Coulter and vanHam forthcoming).

With 13 526 parental home leavers in our researchpopulation, the visualisation technique described abovewould result in a similar number of colour-codedneighbourhood histories. Because of limitations tosoftware, computer screens and printers (limited num-ber of pixels), we had to take a random sample of thesehistories for display in Figures 1–3 (see the figures fordetails of the random samples used). To be certain thatthe figures provide a true representation of the fullpopulation of home leavers, we re-ran the analysesseveral times. The results appeared to be stable (notshown).

Next we modelled the individual neighbourhoodhistories in two different ways. We first modelled theneighbourhood quintile at 5,5 12 and 18 years afterleaving the parental home. The dependent variable isthe neighbourhood quintile based on the percentage oflow-income neighbours. Because the dependent vari-able is ordinal, we used ordered logit regression. Thismodel is also known as the proportional odds modelbecause the odds ratio of the event is independent ofthe relevant category. In ordered logit, a linear functionis estimated of the independent variables and a set ofcut points, which represent an underlying score (for ahousing-related example see Feijten and Mulder 2005;for an example with neighbourhoods see van Ham andManley 2009). The coefficients of the independentvariables can be interpreted as coefficients estimated inan ordinary logit model. Filling in the full regressionequation produces a raw outcome value for eachobservation in the data. To evaluate the probabilitythat an individual lives in one of the five neighbour-hood quintiles it is necessary to calculate a probabilityfor all five categories using the model outcome and thefour cut points (identified as K1 to K4) produced by themodel. For example, the probability to live in a quintile1 neighbourhood is:

P ¼ 11þ expðoutcome� K1Þ

The probability of living in a quintile 2, 3 or 4neighbourhood can be calculated using the followingequation where Kn represents the cut point associatedwith the category in question:

P ¼ 11þ expðoutcome� KnÞ� 11þ expðoutcome� Kn�1Þ

The probability of living in a quintile 5 neighbour-hood is:

P ¼ 1� 11þ expðoutcome� K4Þ

When combined, the probabilities for all five cate-gories will add up to 1, while the most likely destinationneighbourhood will be identified as the one with thehighest probability (see Menard 2002). Finally, we usedlinear regression to model the number of years anindividual home leaver was exposed to quintile 5neighbourhoods (poverty concentration neighbour-hoods) over the 18-year period.

Both the ordered logit models of neighbourhoodquintiles after 5, 12 and 18 years and the linear modelof cumulative exposure to poverty concentration neigh-bourhoods include the same set of independent vari-ables (see Table II for an overview). The mostimportant independent variable is the parental neigh-bourhood quintile in the year before leaving theparental home (1990). Table II shows that in 1995,5 years after leaving the parental home, the distribu-tion by neighbourhood quintile differs significantlyfrom 1990 when children were still living with theirparents. By 1995, the majority of home leavers hadmoved to higher poverty concentration neighbour-hoods (quintile 4 and especially 5). By 2002 manyhome leavers had recovered some of the parentalneighbourhood status and by 2008 the distribution ofneighbourhood poverty status is roughly similar to 1990again.

Household characteristics are measured by twodifferent variables: whether the individual is single orlives in a registered couple (i.e. is married/registeredpartner or is cohabiting with a common child) andwhether the individual has any children below18 years of age. Ethnicity is measured using countryof birth, separating Swedish born from those born inWestern and non-Western countries. In our analyses,we combined the Swedish born and those born inother Western countries and focus especially on thenon-Western born, from here on referred to as‘ethnic minorities’. It was not possible to identifyfirst and subsequent generation immigrants becausedata on parents’ countries of birth were incomplete.The socioeconomic variables include whether theindividual is currently studying, the highest completedlevel of education (where ‘medium’ refers to a high-school degree and a ‘high’ education refers to auniversity degree), income from work (measured in10 000 SEK) and whether the individual receivessocial benefits. Finally, we also control for housingtenure.

In the ordered logit models (Table IV) we havemeasured the independent variables at each of the




modelling years: for 1995 (5 years after leaving theparental home); for 2002 (after 12 years); and for2008 (after 18 years). In the cumulative exposuremodel (Table V) we have recoded some of thevariables so that they measure exposure over the 18-year period (for example, the number of yearssomeone has lived in a public rental dwelling overthe 18-year period). A set of descriptive statistics ofthe data use for the years 1990, 1995, 2002 and 2008can be found in Table II.

Results

Table III shows the cumulative exposure of people to thefive neighbourhood poverty quintiles by parental neigh-

bourhood in 1990. The results show that there is a clearrelationship between the type of parental neighbour-hood and the cumulative exposure to the five neigh-bourhood quintiles over the subsequent neighbourhoodcareer. Those who lived with their parents in a lowpoverty concentration neighbourhood (quintile 1) in1990 are much more likely to spend time in this type ofneighbourhood (17.9% of the next 18 years) than thosewho lived with their parents in a high poverty concen-tration neighbourhood (quintile 5) in 1990 (only 8.9% ofthe next 18 years). Conversely, those who livedwith theirparents in high poverty concentration neighbourhoodsare much more likely to spend time in such a neigh-bourhood (48.8% of the next 18 years) compared withthose who lived with their parents in low poverty

Table II Descriptive statistics of research population in 1990 (when living with parents), 1995, 2002 and 2008

1990 1995 2002 2008

Share males 48.7 48.7 48.7 48.7Age Mean (Std dev.) 21.04 (2.18) 26.04 (2.18) 33.04 (2.18) 39.04 (2.18)Share ethnic minoritiesa 3.2 3.2 3.2 3.2Neighbourhood quintile1 24.7 7.9 17.1 21.22 17.9 8.9 14.6 17.03 15.0 14.6 15.5 18.44 16.2 23.2 22.7 19.55 26.2 45.5 30.1 24.0Share with children 7.4 31.5 63.3 75.4Share couples (ref = singles) 8.2 32.1 59.0 66.7Share students 32.7 14.8 6.5 2.5Level of educationLow 60.3 53.1 45.4 42.6Medium 38.1 40.5 37.6 36.7High 1.6 6.5 17.0 20.7Share employedb – 79.5 89.4 92.1Income from work (10 000 SEK) Mean (Std dev.) 8.89 (5.32) 13.91 (8.13) 24.93 (17.39) 33.94 (29.24)Share receiving social benefits 5.9 6.6 1.5 1.4Housing tenureHome ownership 50.4 12.0 35.8 50.4Cooperative 11.5 33.0 22.4 21.7Private rental 9.2 25.6 16.7 10.5Public rental 19.5 23.8 13.4 10.3N 13 530 13 530 13 530 13 530

Note: Values in percentages unless otherwise stated. Due to missing data on some individuals/years, values do not always add up to 100%aBy ethnic minorities, we refer to people born in non-Western countriesbDue to a change in measurement of employment status between 1992 and 1993, we do not report figures for 1990

Table III Cumulative exposure to neighbourhood income quintiles 1991–2008 (years of exposure as percentage of totalyears)

Parental neighbourhood in 1990 (quintiles)

Cumulative exposure to neighbourhood income quintiles 1991–2008

1 2 3 4 5 Total

1 Low poverty neighbourhood 17.9 14.9 16.0 20.6 30.6 1002 16.3 14.7 16.9 21.9 30.3 1003 13.1 12.8 16.9 23.6 33.6 1004 10.6 10.9 15.7 24.4 38.3 1005 Poverty concentration neighbourhood 8.9 9.0 13.1 20.3 48.8 100

Note: Authors’ calculations on GeoSweden dataset



concentration neighbourhoods (30.6% of the next18 years). These results suggest that the parental neigh-bourhood does indeed have an effect on the neighbour-hood biographies of children during their adult life.Those who grew up in a poverty concentration neigh-bourhood can expect long periods of exposure to thistype of neighbourhood in the rest of their lives. Accord-ing to the neighbourhood effects literature, such expo-sure to poverty concentration neighbourhoods can haveconsequences for individual outcomes.

Figure 1(a)–(c) shows full 18-year neighbourhoodhistories of parental home leavers, organised by theneighbourhood quintile of the parents’ residentialaddress in 1990. Each horizontal line is a uniqueindividual neighbourhood history. A change of colourindicates a move to another neighbourhood quintile.Figure 1(a) shows the neighbourhood histories of thosewhose parents lived in low poverty concentrationneighbourhoods (quintile 1, represented by the colourblue) in 1990. The histories have been ordered basedon neighbourhood quintile in 1990, 1991, 1992, etc.Therefore all individual neighbourhood histories in thisfigure start with a blue line segment. In the first yearafter leaving the parental home, a large group of homeleavers from these relatively affluent neighbourhoodsmove to a poverty concentration neighbourhood (quin-tile 5, represented by the colour grey), but the vastmajority recover neighbourhood status over the sub-sequent years. It is striking to see the variety inneighbourhood histories among our research popula-tion. Previous studies have only investigated year-to-year transitions between neighbourhood types, and weare able to visualise the full histories in all theircomplexity. The final column to the right of the figureshows the same data but sorted by the final destinationneighbourhood quintile in 2008. Here it can be seenthat there is a relatively equal distribution across all 5quintiles, although there seems to be a slight biastowards the higher quintiles. Nevertheless, only a smallproportion of those whose parents lived in the firstquintile end up in the same quintile 18 years later. Anyintergenerational transmission of neighbourhoodadvantage clearly takes a great deal of time to appear.

Figure 1(b) shows the neighbourhood histories ofthose whose parents lived in quintile 3 neighbourhoods(the middle category, represented by the green colour).Figure 1(b) shows a pattern that is roughly comparablewith that in Figure 1(a), although those starting inquintile 3 are slightly more likely to move to quintile 3and 4 neighbourhoods immediately after leaving theparental home. It is striking that those who started inquintiles 1 and 3 have very similar outcomes after18 years (compare the final columns of Figures 1(a)and (b)). After 18 years there is a roughly equaldistribution over the 5 neighbourhood types, regardlessof where people started.

Figure 1(c) shows a radically different picture.These are the histories of those whose parents livedin quintile 5 (high poverty concentration) neighbour-hoods in 1990 (represented by the colour grey).Table III has already shown that these people aremuch more likely than others to be exposed to povertyconcentration neighbourhoods over their life course.Two thirds of the home leavers with parents in a highpoverty concentration neighbourhood move to a similarneighbourhood when they leave the parental home.Over the years, many subsequently move to moreaffluent neighbourhoods. Nevertheless, the final col-umn in Figure 1(c) shows that after 18 years they aremuch more likely than others to live in a povertyconcentration neighbourhood themselves. It is impor-tant to note that the neighbourhood careers of thosestarting in high poverty concentration neighbourhoodsare very diverse. Many histories show episodes inquintile 1 and 2 neighbourhoods (represented by thecolours blue and red), but not as many as in Figure 1(a)with the histories of those starting off in low povertyneighbourhoods. Although there is clear evidence ofintergenerational transmission of neighbourhood pov-erty in Figure 1(a)–(c), the neighbourhood careers ofindividuals starting in similar types of neighbourhoodsare also highly heterogeneous in the short term.

Figure 2 shows the neighbourhood histories ofethnic minority children who lived with their parentsin a poverty concentration neighbourhood (quintile 5)in 1990. The difference with the full population(Figure 1(c)) is striking. Ethnic minorities are muchmore likely than the general population (the majorityof whom are Swedish-born) to move into high povertyconcentration neighbourhood in the year they leave theparental home. They are also much more likely tospend a considerable amount of time in povertyconcentration neighbourhoods during their neighbour-hood histories. However, perhaps the most strikingdifference between Figures 1(c) and 2 is the differencein the final destinations of the ethnic minoritiescompared with the general population. Individualsfrom ethnic minorities with parents in a high povertyconcentration neighbourhood are much more likelythan others (roughly two thirds compared with aboutone third) to end up in a similar type neighbourhoodafter 18 years. The figures demonstrate that neigh-bourhood disadvantage is transmitted particularlystrongly between generations of ethnic minority fami-lies. It is important to keep in mind that the ‘choice’ tolive in low-income neighbourhoods may also be apositive one, based on social networks or proximity tospecific facilities (Manley and van Ham 2011).

Figures 1 and 2 contain selections of neighbourhoodhistories based on the parental neighbourhood in 1990.There are many other ways in which the neighbourhoodhistories can be ordered and categorised. One such




(a)

(b)

(c)

Figure 1 Neighbourhood histories 1990–2008 (10% sample of histories) of those leaving the parental home 1990–1991by parental neighbourhood quintile (1, 3 and 5). (a) Parental neighbourhood quintile 1 in 1990 (low poverty

neighbourhood). (b) Parental neighbourhood quintile 3 in 1990. (c) Parental neighbourhood quintile 5 in 1990 (povertyconcentration neighbourhood)

Source: Authors’ calculations on GeoSweden dataset



alternative categorisation is based on whether peopleshow downward or upward mobility over the period1990–2008, or whether they experience a stable historyover this timeframe. As an illustration, we show thestable mobility histories in Figure 3. In this figure allfive parental neighbourhood quintiles are representedon the left hand side. Each of the histories starts with

the same colour it ends with (the same neighbourhoodquintile) and hence we labelled these stable neigh-bourhood histories. A major advantage of our visuali-sation method is that it reveals that although thehistories are stable in terms of starting and end points,there is a lot of mobility in between. The colour codingclearly shows that the stable quintile 1 histories show

Figure 2 Neighbourhood histories 1990–2008, ethnic minorities (no sample but full population) with parentalneighbourhood quintile 5 in 1990


Figure 3 Stable neighbourhood histories 1990–2008 with same neighbourhood in 2008 as parental neighbourhood in1990 (25% sample)





many more episodes in quintile 1 and 2 (blue and red)than the other histories. On the other hand, the stablequintile 5 (high poverty concentration neighbourhoods)histories contain many more episodes in quintile 4 and5 neighbourhoods. These results show great continuityin neighbourhood status over the life course.

Multivariate modelsTo understand how neighbourhood histories havedeveloped over time, we modelled neighbourhoodoutcomes at 5, 12 and 18 years after leaving theparental home using ordered logit regression(Table IV). This enables us to investigate whether theintergenerational transmission of neighbourhood statusvisible in the figures remains important after control-ling for other attained, inherited or ascribed individualcharacteristics. Informed by the above visualisations,we have included the parental neighbourhood quintilein 1990 and a dummy variable to indicate ethnicity. Tosee if the intergenerational transmission of neighbour-hood poverty is mediated by ethnicity, we included aninteraction effect between ethnicity and parentalneighbourhood type. For each of the years we show amodel with and without parental income in 1990because we want to know whether neighbourhoodoutcomes for children are a result of parental incomeor parental neighbourhood. Since we know from theliterature on the intergenerational transmission ofdisadvantage that the income of parents and childrenis related, it is likely that any similarity in neighbour-hood is simply a result of income. Although this wouldbe an interesting finding in itself, here we are lookingfor an independent effect of the parental neighbour-hood on the neighbourhood outcomes of children.

The results clearly show that the parental neighbour-hood is a strong predictor of neighbourhood (dis)-advantage for children 5, 12 and 18 years after leavingthe parental home. The higher the poverty concentrationof the parental neighbourhood, the higher the povertyconcentration of the neighbourhood of their childrenlater in life. It is important to note that this effect holdsafter controlling for a range of individual and householdcharacteristics, including parental income. The effect ofbeing a non-Western immigrant on neighbourhoodoutcomes is more complicated. The main effect fornon-Western immigrants is not significant in the modelsunless parental income is also included. The onlysignificant effect can be found in themodel with parentalincome at 5 years after leaving the parental home. Here,non-Western immigrants aremuchmore likely to live in apoverty concentration neighbourhood than others. Thisindicates that non-Western immigrants are disadvan-taged in the first years after leaving the parental home,but then catch up later in life. The interaction effectbetween immigrant status and parental neighbourhoodis only significant after 12 and 18 years for those with

parents in the highest poverty concentration neighbour-hoods. However, the effects disappear when controlledfor parental income. This leads to the broad conclusionthat the ethnicity effect found in the visual analysis ofneighbourhood histories is caused by income differencesbetween groups.

The control variables show that there are nosignificant gender or child effects on neighbourhoodoutcomes. Those living with a partner are less likely toend up in poverty concentration neighbourhoods thansingles. This is most likely due to the higher level ofresources available to households with two earners.After 5 years, those with a middle level of educationare the least likely to end up in poverty concentrationneighbourhoods compared with those with lower anduniversity-level education. For those with a university-level education, this can be explained by the fact thatthey start their housing career somewhat later due toinvestments in their human capital. This is confirmedby the finding that after 18 years those with a universitydegree are the least likely to end up in a povertyconcentration neighbourhood. Greater levels of incomefrom work reduce the probability of ending up in apoverty concentration neighbourhood, while being onsocial benefits increases the probability. Those living inpublic rented accommodation are the most likely toend up in poverty concentration neighbourhoods,followed by those in private renting, cooperativehousing and owner-occupied housing.

The final models in Table V report the effect ofcumulative exposure to high poverty concentrationneighbourhoods (quintile 5) over the full 18-yearperiod after leaving the parental home. The maximumexposure time in this model is therefore 18 years. Theresults clearly demonstrate that individuals who livedwith their parents in quintile 4 and especially quintile 5neighbourhoods in 1990 spend significantly longer inpoverty concentration neighbourhoods over the next18 years than those who grew up in the low povertyconcentration neighbourhoods. Non-Western immi-grants have especially long exposure times to highpoverty concentration neighbourhoods, even aftercontrolling for parental income. The interaction effectsbetween immigrant status and parental neighbourhooddo not indicate an additional effect for immigrants(which was the case in a model without parentalincome, effects not shown). The control variables showthat having a middle-level income reduces the cumu-lative exposure to poverty concentration neighbour-hoods. Having a high mean income and an increase inincome during the 18 years (measured by incomerange) also reduce cumulative exposure to povertyneighbourhoods. In contrast, receiving social benefitsincreases exposure to poverty concentration neighbour-hoods. With increasing number of years in publicrenting, the exposure to poverty concentration neigh-



TableIV

Ordered

Logitmod

elsof

neighb

ourhoo

dqu

intile

(1–5)5,

12an

d18

yearsafterleavingthepa

rental

home

After

5yearsa

After

12years

After

18years

Coe

ff.

StdErr

Coe

ff.

StdErr

Coe

ff.

StdErr

Coe

ff.

StdErr

Coe

ff.

StdErr

Coe

ff.

StdErr

Parent

NBH

Q1

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Parent

NBH

Q2

�0.004

0.052

�0.002

0.053

0.083

0.050

0.080

0.052

0.106*

0.049

0.125*

0.051

Parent

NBH

Q3

0.117*

0.055

0.113*

0.057

0.249***

0.053

0.244***

0.055

0.189***

0.052

0.197***

0.054

Parent

NBH

Q4

0.267***

0.054

0.249***

0.056

0.438***

0.053

0.440***

0.055

0.349***

0.052

0.340***

0.054

Parent

NBH

Q5

0.395***

0.049

0.405***

0.052

0.506***

0.048

0.523***

0.050

0.363***

0.047

0.387***

0.049

Non

�Western

immigrant

0.501

0.317

1.646*

0.724

�0.401

0.297

0.249

0.590

�0.078

0.296

0.488

0.593

Parent

NBH

Q1*ethn

icminority

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Parent

NBH

Q2*ethn

icminority

�1.017*

0.431

�2.235**

0.848

0.418

0.425

0.474

0.768

�0.123

0.405

�0.723

0.734

Parent

NBH

Q3*ethn

icminority

0.040

0.472

�0.394

1.006

1.311**

0.456

0.657

0.834

0.625

0.436

0.062

0.831

Parent

NBH

Q4*ethn

icminority

�0.374

0.434

�0.998

0.980

1.010*

0.417

0.805

0.826

0.319

0.398

0.230

0.802

Parent

NBH

Q5*ethn

icminority

0.333

0.359

�0.445

0.765

1.196***

0.333

0.819

0.623

1.080**

0.332

0.697

0.626

Female

�0.034

0.036

�0.039

0.038

0.042

0.036

0.054

0.038

0.013

0.035

0.038

0.037

Children

0.061

0.063

0.062

0.066

�0.075

0.054

�0.090

0.057

0.090

0.054

0.057

0.057

Cou

ple

�0.164**

0.061

�0.176**

0.064

�0.106*

0.054

�0.098

0.057

�0.173**

0.050

�0.181**

0.053

Low

education

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Middleed

ucation

�0.116**

0.037

�0.120**

0.040

�0.024

0.037

�0.045

0.040

�0.169***

0.037

�0.175***

0.039

University

degree

�0.064

0.070

�0.070

0.075

0.105*

0.048

0.073

0.052

�0.207***

0.045

�0.210***

0.048

Stud

ent

�0.038

0.056

�0.026

0.060

0.006

0.071

�0.009

0.076

�0.016

0.103

�0.026

0.111

Incomefrom

work

(10000SE

K)

�0.014***

0.003

�0.014***

0.003

�0.001

0.001

� 0.001

0.001

�0.003***

0.001

�0.002**

0.001

Social

bene

fits

0.116

0.081

0.092

0.087

0.411*

0.159

0.299

0.173

0.619***

0.169

0.541**

0.181

Hom

eow

nership

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Ref

Coo

pho

using

1.072***

0.056

1.068***

0.059

1.473***

0.045

1.455***

0.048

1.158***

0.043

1.130***

0.045

Privaterenting

1.573***

0.060

1.557***

0.063

2.154***

0.052

2.168***

0.055

1.785***

0.057

1.744***

0.061

Public

renting

2.607***

0.065

2.549***

0.068

2.898***

0.061

2.889***

0.065

2.921***

0.067

2.913***

0.072

Parent

incomefrom

work

(10000SE

K)

0.000

0.001

0.004*

0.001

0.001

0.001

/cut1

�1.458

0.083

�1.489

0.091

�0.628

0.065

�0.553

0.074

�0.821

0.063

�0.792

0.071

/cut2

�0.516

0.081

�0.550

0.089

0.379

0.065

0.463

0.074

0.140

0.063

0.161

0.071

/cut3

0.420

0.081

0.389

0.089

1.256

0.066

1.337

0.075

1.0842

0.064

1.107

0.072

/cut4

1.578

0.082

1.550

0.090

2.541

0.069

2.629

0.078

2.235

0.066

2.266

0.074

Initial

LL

�17692

�15932

�19928

�17928

�20845

�18708

Final

LL

�16392

�14803

�17563

�15833

�18826

�16959

Num

berof

observations

b12

743

11421

12686

11373

13004

11663

Prob

>chi2

0.0000

0.0000

0.0000

0.0000

0.0000

0.0000

Pseu

doR

20.0735

0.0709

0.1187

0.1168

0.0968

0.0935

a Weuse5insteadof

6yearshe

redu

eto

therelia

bilityof

theho

usingda

taavailablein

year

6.b T

hedifferen

cesin

thenu

mbe

rof

observations

betweenmod

elsaredu

eto

smallam

ountsof

missing

data.

*p<0.10;**p<0.05;***p

<0.01




bourhoods increases, while spending greater periods oftime in home ownership reduces the exposure to themost disadvantaged neighbourhoods.

Conclusions

This is the first study to construct and visualise theentire neighbourhood histories of a large group ofindividuals over a long period of time. We investigatedthe intergenerational transmission of neighbourhoodpoverty in Stockholm through the effect of the parentalneighbourhood on individual neighbourhood historiesover a period of almost two decades. By constructingneighbourhood histories, this study has sought toempirically operationalise the concept of unique indi-vidual biographies emphasised by life course theory(Dykstra and van Wissen 1999). We argued thataccurately measuring the extent to which parentalneighbourhood context is transmitted to children and

understanding the factors that lead to neighbourhoodsorting by individuals is critical to understandingresidential outcomes later in life. There is a vastliterature on neighbourhood effects that ties individualoutcomes to the neighbourhood in which they currentlylive. By taking a much longer term view we havedemonstrated that individual outcomes are influencedover a much longer timescale: where individuals livedup to 18 years ago is important for their currentoutcomes.

Using innovative visualisation techniques, we haveshown that individuals sort themselves into neighbour-hoods across the income spectrum as they movethrough the life course. This sorting process is theresult of individual preferences, resources, but alsoopportunities and constraints offered by the housingmarket. The graphs clearly showed that although manyindividuals experienced an initial drop in neighbour-hood status immediately after leaving the parentalhome, many catch up in their subsequent residentialcareer. However, we also demonstrated that childrenliving with their parents in high poverty concentrationneighbourhoods are very likely to end up in similarneighbourhoods much later in life when they are adults,especially when they are an ethnic minority. Theseresults were confirmed by the multivariate analyses.The results show that the intergenerational transmis-sion of disadvantage is a powerful mechanism explain-ing the residential outcomes of individuals across theirlife course. It is important to note that we found theseresults for Sweden, one of the Nordic countries morecommonly associated with equality in outcomes. Basedon our results we would expect to find even strongerintergenerational transmission of disadvantage in acountry like the UK, which has a more segmentedhousing market and a more unequal income distribu-tion than Sweden. The study has also shown that theparental neighbourhood is highly predictive for thecumulative exposure to poverty concentration neigh-bourhoods over the life course. Those individuals wholived with their parents in high poverty concentrationneighbourhoods spend significantly longer in suchneighbourhoods over the next 18 years than thosewho grew up in low poverty concentration neighbour-hoods. Ethnic minorities were found to have thelongest cumulative exposure to poverty concentrationneighbourhoods.

That parental neighbourhood type has such a long-lasting impact on exposure, even after controlling for avariety of changes occurring elsewhere in the lifecourse, suggests that disadvantage is not solely trans-mitted through parental income, but is also linked toliving in poverty neighbourhoods. Based on our data wecannot identify the causal mechanisms underlying theintergenerational transmission of neighbourhood pov-erty. There are several possible explanations for the

Table V Linear regression of years of exposure (min 0and max 18 years) to poverty concentration (quintile 5)neighbourhoods after leaving the parental home

Coeff. Std. Err.

Parent NBH Q1 Ref RefParent NBH Q2 �0.086 0.130Parent NBH Q3 0.279 0.139**Parent NBH Q4 0.683 0.136***Parent NBH Q5 1.586 0.124***Non–Western immigrant 3.395 1.412***Parent NBH Q1*ethnic minority Ref RefParent NBH Q2*ethnic minority �2.610 1.794Parent NBH Q3*ethnic minority �0.586 1.920Parent NBH Q4*ethnic minority �0.084 1.836Parent NBH Q5*ethnic minority �0.100 1.463Female �0.129 0.095Number of years withchildren (0–18)

0.006 0.009

Number of years in couple (0–18) �0.008 0.010Low education Ref RefMiddle education �0.272 0.104***University degree �0.243 0.163Number of years studying (0–18) �0.011 0.025Mean income from work(10 000 SEK)

�0.046 0.006***

Work income range 0.010 0.002***Number of years on socialbenefits (0–18)

0.134 0.025***

Number of years in publicrental (0–18)

0.316 0.009***

Number of years in homeownership (0–18)

�0.207 0.009***

Parent income from work(10 000 SEK)

0.003 0.004

Constant 7.019 0.202***Number of obervationsa 12 105F 230.59Adjusted R2 0.2849

aThe number of observations < total sample due to missing data.*p < 0.10; **p < 0.05; ***p < 0.01



effects found. One such explanation is that childrenchoose to live in neighbourhood types that are familiarfrom their childhood. Another possible explanationcould be that children choose to live in close proximityto their parents, and therefore end up in similarneighbourhoods, or choose to live in a neighbourhoodwith similar (ethnic group) specific services. The effectsmight also be the result of a lack of choice due to thestructure and institutional arrangements of the housingmarket. A more complex set of explanations might bederived from the neighbourhood effects literature.Socialisation and peer group effects, and a lack ofpositive role models (see Galster 2012), might leadthose growing up in high poverty concentration neigh-bourhoods to develop deviating labour market attitudesand values, and might lead to a lack of informal skillsneeded on the labour market. Although we control forindividual income in our models, income might not be asufficient proxy for these factors which in turn mightaffect the choice (or lack of choice) to live in a povertyconcentration neighbourhood as adults. Futureresearch can shed more light on these possible under-lying mechanisms.

This study has showcased the contribution thatgeography as an academic discipline can make to thediscussion on intergenerational mobility. Where soci-ologists focus mainly on intergenerational mobilityacross class and occupations, and economists typicallyanalyse income and earnings mobility (D’Addio 2007;Solon 1999), our study has demonstrated the impor-tance of the spatial dimension of intergenerationaltransmission of (dis)advantage. The results of our studyalso have wider implications for other areas ofgeographical study. The literature on neighbourhoodeffects typically uses simple point-in-time measures ofneighbourhood. We have demonstrated that it isimportant to measure neighbourhood experiences overlonger periods of time because for some groupsdisadvantage is both inherited and highly persistent,while for other groups living in poverty neighbourhoodsis only a temporary state. We have clearly demon-strated that adult exposure to poverty concentrationneighbourhoods is linked to the neighbourhood that anindividual lived in with their parents. This indicates thatneighbourhood effects might run between generationsand suggests that not just the current neighbourhood,but the whole neighbourhood history, should be takeninto account when investigating whether people aredisadvantaged by where they live. The results also havewider implications for the study of segregation. Under-standing why people live in certain types of neighbour-hoods cannot be simply understood from their currentcharacteristics, but a full understanding should takeinto account the neighbourhood biographies of indi-viduals. Finally, this study has demonstrated the valueof visualisation techniques for spatial and temporal

data. Visualising neighbourhood histories allowed us tocontextualise events and transitions within a true lifecourse framework.

Acknowledgements

The research reported in this paper was made possiblethrough the financial support of the Institute forHousing and Urban Research (IBF) at UppsalaUniversity, Sweden and the financial support of theEU (NBHCHOICE Career Integration Grant underFP7-PEOPLE-2011-CIG). We would like to thank theeditor and the anonymous reviewers for their com-ments.

Notes

1 The Stockholm metropolitan region includes the munici-palities of Stockholm and Solna, along with those munic-ipalities of the Stockholm labour market region where themajority of the commuting flow is into either Stockholm orSolna.

2 The calculations stop when the number of neighboursexceeds 500. Since the software uses a 100 by 100 metregrid, the total number of neighbours included is oftenslightly higher than 500.

3 Income from work is calculated as the sum of: salarypayments, income from active businesses and tax-basedbenefits that employees accrue as terms of their employ-ment (including sick or parental leave, work-related injuryor illness compensation, daily payments for temporarymilitary service or giving assistance to a disabled relative).

4 The figures and tables that follow are presented as near aspossible to the primary discussion of the relevantinformation, rather than necessarily in the order in whichthey are first mentioned.

5 The reason for modelling outcomes after 5 (in 1995)instead of the more logical 6 years is that data on tenureonly is available in 1990, 1995, 2000, 2002, 2004, 2006 and2008. We have roughly estimated tenure for interveningyears.

References

Aisenbrey S and Fasang A E 2010 New life for old ideas: the‘second wave’ of sequence analysis bringing the ‘course’ backinto the life course Sociological Methods and Research 38420–62

Becker G S and Tomes N 1979 An equilibrium theory of thedistribution of income and intergenerational mobility Jour-nal of Political Economy 87 1153–89

Bisin A and Verdier T 1998 On the cultural transmission ofpreferences for social status Journal of Public Economics 7075–97

Blaauboer M 2011 The impact of childhood experiences andfamily members outside the household on residentialenvironment choices Urban Studies 48 1635–50

Brzinsky-Fay C, Kohler U and Luniak M 2006 Sequenceanalysis with Stata The Stata Journal 6 435–60




Cabinet Office 2011 Opening doors, breaking barriers: a strategyfor social mobility Cabinet Office, London

Clark W A V and Withers S D 2007 Family migration andmobility sequences in the United States: spatial mobility inthe context of the life course Demographic Research 17 591–622

Clark W A V and Dieleman F M 1996 Households and housing:choice and outcomes in the housing market CUPR Press, NewBrunswick NJ

Clark W A V and Huang Y 2003 The life course and residentialmobility in British housing markets Environment and Plan-ning A 35 323–39

Clark W A V and Ledwith V 2006 Mobility, housing stress, andneighborhood contexts: evidence from Los Angeles Envi-ronment and Planning A 38 1077–93

Clark W A V, Deurloo M and Dieleman F M 2003 Housingcareers in the United States, 1968–93: modelling thesequencing of housing states Urban Studies 40 143–60

Coulter R and van Ham M forthcoming Following peoplethrough time: an analysis of individual residential mobilitybiographies Housing Studies

Crowder K and South S J 2005 Race, class, and changingpatterns of migration between poor and nonpoor neighbor-hoods American Journal of Sociology 110 1715–63

D’Addio A C 2007 Intergenerational transmission of disadvan-tage: mobility or immobility across generations OECD Social,Employment and Migration Working Papers No. 52 OECDPublishing, Paris

Duncan G J and Raudenbush S W 2001 Neighborhoods andadolescent development: how can we determine the links? inBooth A and Crouter A C eds Does it take a village?Community effects on children, adolescents, and familiesLawrence Erlbaum, Mahwah NJ 105–36

Durlauf S N 2004 Neighbourhood effects in Henderson J Vand Thisse J F eds Handbook of regional and urbaneconomics. Volume 4: Cities and geography Elsevier, Amster-dam 2173–242

Dykstra P and van Wissen L 1999 Introduction: the life courseapproach as an interdisciplinary framework for populationstudies in van Wissen L and Dykstra P eds Population issues:an interdisciplinary focus Plenum Press, New York 1–22

Elder G H 1994 Time, human agency, and social change:perspectives on the life course Social Psychology Quarterly 574–15

Ellen I G and Turner M A 1997 Does neighbourhoodmatter? Assessing recent evidence Housing Policy Debate 8833–66

Esping-Andersen G 1990 The three worlds of welfare capitalismPrinceton University Press, Princeton NJ

Feijten P 2005 Life events and the housing career: a retrospectiveanalysis of timed effects Eburon, Delft

Feijten P and Mulder C H 2005 Life course experience andhousing quality Housing Studies 20 571–88

Feijten P, Hooimeijer P and Mulder C H 2008 Residentialexperience and residential environment choice over thelife-course Urban Studies 45 141–62

Galster G 2002 An economic efficiency analysis of deconcen-trating poverty populations Journal of Housing Economics 11303–29

Galster G 2012 The mechanism(s) of neighbourhood effects:theory, evidence, and policy implications in van Ham M,

Manley D, Simpson L, Bailey N and Maclennan D edsNeighbourhood effect research: new perspectives Springer,Dordrecht 23–56

Geist C and McManus P A 2008 Geographical mobility overthe life course: motivations and implications Population,Space and Place 14 283–303

Hedin K, Clark E, Lundholm E and Malmberg G 2012Neoliberalization of housing in Sweden: gentrification,filtering, and social polarization Annals of the Associationof American Geographers 102 443–63

Helderman A and Mulder C H 2007 Intergenerationaltransmission of homeownership: the roles of gifts andcontinuities in housing market characteristics Urban Studies44 231–47

Henretta J C 1984 Parental status and child’s home ownershipAmerican Sociological Review 49 131–40

Jencks C and Mayer S E 1990 The social consequences ofgrowing up in a poor neighborhood in Lynn L E andMcGeary M G H eds Inner-city poverty in the United StatesNational Academy Press, Washington DC 111–86

Kasarda J D 1988 Jobs, migration, and emerging urbanmismatches in McGeary M G H and Lynn L E eds Urbanchange and poverty National Academy Press, Washington DC148–98

Kim T K, Horner M W and Marans R W 2005 Life cycle andenvironmental factors in selecting residential and joblocations Housing Studies 20 457–73

Kuntz J, Page M E and Solon G 2003 Are point-in-timemeasures of neighborhood characteristics useful proxies forchildren’s long-run neighborhood environment? EconomicLetters 79 231–7

Kurz K 2004 Labour market position, intergenerational trans-fers and home-ownership. A longitudinal analysis for WestGerman birth cohorts European Sociological Review 20 141–59

Maloutas T 2012 Introduction: residential segregation incontext in Maloutas T and Fujita K eds Residential segrega-tion in comparative perspective. Making sense of contextualdiversity Ashgate, Farnham 1–36

ManleyDJ and vanHamM 2011Choice-based letting, ethnicityand segregation in England Urban Studies 48 3125–43

Manley D J and van Ham M 2012 Neighbourhood effects,housing tenure, and individual employment outcomes in vanHam M, Manley D, Bailey N, Simpson L and Maclennan Deds neighbourhood effects research: new perspectives Springer,Berlin 147–74

Massey D S, Gross A B and Shibuya K 1994 Migration,segregation, and the concentration of poverty AmericanSociological Review 59 425–45

Meen G, Meen J and Nygaard C 2007 A tale of two Victoriancities in the 21st century ICHUE – Discussion papers No. 007

Menard S 2002 Applied logistic regression analysis Sage, LondonMulder C H and Hooimeijer P 1999 Residential relocations inthe life course in van Wissen L and Dykstra P eds Populationissues: an interdisciplinary focus Plenum Press, New York159–86

Mulder C H and Smits J 1998 First-time home-ownership ofcouples. The effect of inter-generational transmission Euro-pean Sociological Review 15 323–37

OECD 2010 Economic policy reforms 2010: going for growthOrganization for Economic Cooperation and Development,Paris



Östh J, Malmberg B and Andersson E forthcoming Analysingsegregation using individualized neighbourhoods in Lloyd CD, Shuttleworth I and Wong D eds Social-spatial segregation:concepts, processes and outcomes Policy Press, Bristol

Pollock G 2007 Holistic trajectories: a study of combinedemployment, housing and family careers by using multi-ple-sequence analysis Journal of the Royal Statistical Society:Series A 170 167–83

Quillian L 2003 How long are exposures to poor neighbor-hoods? The long-term dynamics of entry and exit from poorneighbourhoods Population Research and Policy Review 22221–49

Rabe B and Taylor M 2010 Residential mobility, quality ofneighbourhood and life course events Journal of the RoyalStatistical Society: Series A 173 531–55

Rossi P H 1980 Why families move. A study in the socialpsychology of urban residential mobility Free Press, Glencoe IL

Sampson R J and Wilson W J 1995 Toward a theory of race,crime, and urban inequality inHagan J and Peterson R D edsCrime and Inequality Stanford University Press, Stanford, CA177–90

Simpson L and Finney N 2009 Spatial patterns of internalmigration: evidence for ethnic groups in Britain Population,Space and Place 15 37–56

Solon G 1999 Intergenerational mobility in the labor market inAshenfelter O, Layard R and Card D eds Handbook of laboreconomics volume 3 Elsevier, Dordrecht 1761–800

South S J and Crowder K D 1997 Escaping distressedneighborhoods: individual, community, and metropolitaninfluences American Journal of Sociology 102 1040–84

South S J and Deane G D 1993 Race and residential mobility:individual determinants and structural constraints SocialForces 72 147–67

Stovel K and Bolan M 2004 Residential trajectories: usingoptimal alignment to reveal the structure of residentialmobility Sociological Methods and Research 32 559–98

van Ham M 2012 Housing behaviour in Clapham D, Clark WA V and Gibb K eds Handbook of housing studies Sage,London chapter 3

van Ham M and Manley D 2009 Social housing allocation,choice and neighbourhood ethnic mix in England Journal ofHousing and the Built Environment 24 407–22

van Ham M and Manley D 2010 The effect of neighbourhoodhousing tenure mix on labour market outcomes: a longitu-dinal investigation of neighbourhood effects Journal ofEconomic Geography 10 257–82

van Ham M, Manley D, Bailey N, Simpson L and MaclennanD 2012 eds Neighbourhood effects research: new perspectivesSpringer, Dordrecht

Vartanian T P, Buck P W and Gleason P 2007 Intergenera-tional neighborhood-type mobility: examining differencesbetween blacks and whites Housing Studies 22 833–56

Wilson W J 1987 The truly disadvantaged University of ChicagoPress, Chicago IL




Intergenerational transmission of neighbourhood poverty: an analysis of neighbourhood histories of...

Documents

Transcript of Intergenerational transmission of neighbourhood poverty: an analysis of neighbourhood histories of...