Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Income polarization and the emergence of a low income SkyTrain corridor in Metro Vancouver, 1971-2006 Jones, Craig E. 2014

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


24-ubc_2014_november_jones_craig.pdf [ 1.79MB ]
JSON: 24-1.0167002.json
JSON-LD: 24-1.0167002-ld.json
RDF/XML (Pretty): 24-1.0167002-rdf.xml
RDF/JSON: 24-1.0167002-rdf.json
Turtle: 24-1.0167002-turtle.txt
N-Triples: 24-1.0167002-rdf-ntriples.txt
Original Record: 24-1.0167002-source.json
Full Text

Full Text

INCOME POLARIZATION AND THE EMERGENCE OF A LOW INCOMESKYTRAIN CORRIDOR IN METRO VANCOUVER, 1971-2006byCraig E. JonesB.A., The University of British Columbia, 2012A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF ARTSinTHE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES(Geography)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)September 2014© Craig E. Jones , 2014iiAbstractIncome inequality is on the increase internationally, in Western Anglophone nations, and inCanadian cities. In Metro Vancouver, broad processes of socio-spatial polarization in the regionhave led to the emergence of a low-income corridor that follows the SkyTrain Expo Line rapidtransit alignment from East Vancouver to North Surrey. Within this low-income corridor thereare processes of significant and varied neighbourhood change. The population of MetroVancouver continues to expand, with this growth largely fueled by immigration. In order toaccommodate population growth, increased residential density and transit-oriented developmentnear SkyTrain stations has become a common public policy prescription.A mixed-methodology was pursued in order to conduct research at three geographical scales. Atthe scale of the Vancouver region, quantitative analysis is employed to discover associationsbetween the demographics and the housing stock of neighbourhoods, and the average incomes ofcensus tracts as well as changes in average incomes. At the sub-regional scale of the low-incomecorridor, semi-structured interviews were conducted with 14 key informants. Finally, at the localscale, two neighbourhoods were selected for further study. Four focus groups were conductedwith a total of 26 residents of two neighbourhoods within the low-income corridor.The key findings of these methods are generally complementary. At the regional scale, theVancouver CMA has seen an increase of income inequality. The socio-spatial polarization ofneighbourhoods is a consequence of income inequality and these processes are stronglyassociated with visible minority status, immigration, and apartment unit dwellings. These threefactors are central to understanding the emergence of, and dynamics in, the low-income corridor.iiiPolicies which encourage high-density development near SkyTrain stations in the low-incomecorridor have increased development pressure; particularly so for one district of affordable rentalapartments which is largely occupied by recent immigrants of visible minority status.I urge scholars conducting research in urban income inequality to incorporate mixed methodsand multi-scalar analysis into their research design. In doing so, our findings will be enriched,textured and challenged, and our projects made stronger.ivPrefaceThe design of the research program, all primary and secondary research, and all analysis of datain this thesis was conducted exclusively by Craig E. Jones. For Chapter 3 a series of sixmultivariate regressions were conducted using data from the 1996 and 2006 Canadian censuses.For Chapter 4 semi-structured interviews were conducted with fourteen key informants. All keyinformants were contacted via email by Craig E. Jones, and all interviews were conducted byCraig E. Jones. For Chapter 5 four focus groups were conducted in two neighbourhoods with atotal of 26 participants. Participants were recruited by two third-party organizations, and thefocus groups were conducted by Craig E. Jones. All interviews and focus groups weretranscribed and analyzed by Craig E. Jones.The field-work for this thesis was granted ethics approval by the UBC Behavioural ResearchEthics Board under certificates: H13-01282, H13-03225.vTable of ContentsAbstract.......................................................................................................................................... iiPreface........................................................................................................................................... ivTable of Contents ...........................................................................................................................vList of Tables .............................................................................................................................. viiiList of Figures............................................................................................................................... ixList of Abbreviations .....................................................................................................................xAcknowledgements ...................................................................................................................... xiChapter 1: Introduction ................................................................................................................1Chapter 2: Literature Review.......................................................................................................52.1 Theoretical Frameworks ................................................................................................. 52.2 Income Inequality at the International and National Scales ........................................... 62.3 Income Inequality in Cities ............................................................................................. 82.4 Income Inequality at the Neighbourhood Scale............................................................ 122.5 Methods and Data ......................................................................................................... 14Chapter 3: Quantitative Study of Income Inequality in the Vancouver CMA......................163.1 Selection of Independent Variables .............................................................................. 173.1.1 The Independent Variables for 2006......................................................................... 183.2 Regression Series 1: OLS Multivariate Regressions .................................................... 193.2.1 Dependent Variables of the OLS Regressions of Average Individual and HouseholdIncomes, 2006 ....................................................................................................................... 213.2.2 Results of OLS Multivariate Regressions................................................................. 21vi3.2.3 Discussion of OLS Multivariate Regression Results................................................ 213.3 Regression Series 2: Logistic Regressions Using 2006 Census Data........................... 233.3.1 Dependent Variable for Logistic Regression of Individual Income Decline............ 263.3.2 Dependent Variable for Logistic Regression of Household Income Decline........... 273.3.3 Results of Multivariate Logistic Regressions of Income Decline ............................ 283.3.4 Discussion of Multivariate Logistic Regressions Results......................................... 283.4 Regression Series 3: Logistic Regressions of Change, 1996-2006............................... 313.4.1 Results from Multivariate Logistic Regressions of Change, 1996-2006.................. 343.5 Conclusion .................................................................................................................... 37Chapter 4: The Emerging SkyTrain Low-Income Corridor ...................................................424.1 Origins of the Corridor.................................................................................................. 474.2 Rental Housing and Immigration.................................................................................. 494.3 A Corridor of Contingent Development ....................................................................... 554.3.1 Vancouver ................................................................................................................. 564.3.2 Burnaby..................................................................................................................... 594.3.3 New Westminster...................................................................................................... 624.3.4 Surrey........................................................................................................................ 634.4 Issues Associated with Development: Land Speculation and Displacement................ 654.5 Conclusion .................................................................................................................... 69Chapter 5: Neighbourhood Change in the Low-Income Corridor .........................................735.1 Neighbourhood Selection Process ................................................................................ 745.2 Community Profiles of Maywood and Richmond Park................................................ 795.3 Focus Groups in Maywood and Richmond Park .......................................................... 81vii5.3.1 Recruitment............................................................................................................... 815.3.2 Selected Demographic Characteristics of Focus Group Participants ....................... 825.3.3 Major Insight from Focus Groups............................................................................. 825.3.4 Perception of Development in Richmond Park ........................................................ 885.3.5 Perception of Development in Maywood ................................................................. 915.4 Understanding the Different Perceptions of Development ........................................... 945.5 Conclusion .................................................................................................................. 101Chapter 6: Conclusion...............................................................................................................105Bibliography ...............................................................................................................................114viiiList of TablesTable 1 Results of OLS Multivariate Regressions........................................................................ 21Table 2 Results of Multivariate Logistic Regressions of Individual and Household IncomeDecline .......................................................................................................................................... 28Table 3 Results from Multivariate Logistic Regressions of Change, 1996-2006......................... 34Table 4 Community Profiles of Maywood and Richmond Park................................................... 79Table 5 Selected Demographic Characteristics of Focus Group Participants .............................. 82Table 6 ‘s’ Zoning and Maximum FAR ....................................................................................... 95ixList of FiguresFigure 1 Change in Census Tract Average Individual Income, 1970-2005 ................................. 24Figure 2 Change in Census Tract Average Household Income, 1970-2005................................. 24Figure 3 Census Tract Average Individual Income, 1971 ............................................................ 43Figure 4 Census Tract Average Individual Income, 2006 ............................................................ 44Figure 5 Census Tract Average Household Income, 1971 ........................................................... 45Figure 6 Cesus Tract Average Household Income, 2006 ............................................................. 45Figure 7 Route of the Inter-Urban Line ........................................................................................ 47Figure 8 Recent Immigrants in Vancouver CMA......................................................................... 51Figure 9 Metro Vancouver GAR Arrivals, 2005-2009................................................................. 53Figure 10 Metro Vancouver GAR Arrivals, 2010-2013............................................................... 54Figure 11 Refugees in Vancouver CMA ...................................................................................... 55Figure 12 Very Low Average Household Income in Maywood and Richmond Park.................. 75Figure 13 Recent Immigrants in Maywood and Richmond Park ................................................. 76Figure 14 Refugees in Maywood and Richmond Park ................................................................. 77Figure 15 Recent Refugees in Maywood and Richmond Park ..................................................... 78Figure 16 Rental Housing Disadvantage in Maywood and Richmond Park ................................ 79Figure 17 Apartment Demolition Permits in Burnaby, 2002-2014 .............................................. 96xList of AbbreviationsCMA: census metropolitan areaCT: census tractFAR: floor-area ratioGAR: government assisted refugeeISSofBC: Immigrant Services Society of BCNHS: National Household SurveyOLS: ordinary least squaresPBRH: purpose-built rental housingSBTC: skills-biased technological changeSEG: socio-economic groupxiAcknowledgementsIt would not have been possible to complete this thesis without extensive the network of supportthat I am incredibly fortunate to have. I would like to thank my supervisor, Dr. David Ley forgenerously sharing your knowledge, wisdom, and insight. Without your support this projectwould not have happened. Thank you to my second reader, Dr. Daniel Hiebert for enriching myresearch practice and for providing me with opportunities to move towards ever-greater rigour.I am grateful to the Vancouver group of the Neighbourhood Change Research Partnership andthe UBC Faculty of Graduate and Postdoctoral Studies for their financial support of this project.I would like to extend my thanks to the long list of graduate students in the Department ofGeography at UBC who were always willing to give their advice and share their experiencesover the last two years. Special thanks to Paige Patchin for your invaluably constructivefeedback.Thank you, Naida Slakov for your love, understanding and support.To Dr. Elvin Wyly, your contribution to this project started long before it began and continued tothe very end. My gratitude exceeds the boundaries of the quantifiable.Thank you to the staff of South Burnaby Neighbourhood House and MOSAIC BC for yoursupport of this project, and thank you to all of the key informants and focus groups participantswho agreed to share their valuable time and expertise.1Chapter 1: IntroductionRecent research into urban income inequality has revealed a transition in Canadian cities over thelast 40 years (see Hulchanski et al. 2007, Brzozowski 2010, Ades et al. 2012, Chen et al. 2012,Ley & Lynch 2012). Traditional inner-city areas of poverty have become revalorized in post-industrial cities while some middle-income suburban districts have transitioned into low-incomeareas, reversing our traditional understanding of the geography of urban income distributions.Within this context of transition and polarization in Greater Vancouver, a distinctive zone ofdistricts along the rapid transit Skytrain corridor has moved into low income status since 1971, atrend that has accelerated since 2000. This thesis identifies some of the factors associated withsocio-spatial polarization in the Vancouver census metropolitan area (CMA) which have led tothe development of a low-income (and in several neighbourhoods, very-low income) region informerly middle-income neighbourhoods, straddling four municipalities from inner cityVancouver through to the suburbs of (South) Burnaby, New Westminster and (North) Surrey.This thesis has been guided by a series of six research questions:1. What factors, such as demographics and housing characteristics are associated withdifferences in the average income of census tracts (CT) and changes in the averageincome of CTs over time?2. Why has this corridor of low-incomes developed, and what does the location provide itspopulation?3. What are the demographic characteristics of households along this corridor?4. What share of the corridor’s population is new immigrants and have co-ethnic clustersformed?5. What has been the role of public policy in the development of this low-income region?6. How might future policies – e.g. the redevelopment of sites near Skytrain stations - affect2the stability of affordable housing in this district?This research has been informed by and builds upon the work of Hulchanksi et al. (2007) whichhas generated interest in socio-spatial polarization in Canadian cities, and Ley & Lynch (2012)which provides the guiding framework for this thesis. The findings of Ley & Lynch (2012) areexpanded upon through a mixed-methods approach at three geographical scales.1. Regional scale: quantitative analysis of Vancouver CMA and the low-income corridor,drawing upon the 1996, 2006, and 2011 censuses.2. Sub-regional scale: Semi-structured interviews with 14 professionals and communityleaders.3. Neighbourhood scale: four focus groups in two neighbourhoods with a total of 26participants, drawn from service agency lists.4. Supplemental sources: review of newspaper accounts, planning reports and other publicdocuments.This thesis is organized as follows. Chapter 2 serves as a literature review of recent scholarshipon income inequality. Income inequality in developed nations has increased drastically in the last40 years (see Piketty & Saez 2006) and there are two broad theoretical frameworks which areprominent in explaining income polarization; skills-biased technical change (Bound & Johnson1992; Berman et al. 1994; Katz & Murphy 1992; Krueger 1993; Levy & Murnane 1992; Juhn etal. 1993) and the world and global cities hypothesis (Friedman 1986; Sassen 1988). Thesetheoretical frameworks are either challenged or supported by empirical studies which employseveral methods at a variety of scales. Income inequality has increased at the international scale(Piketty & Saez 2006) and at the national scale of the US (Autor et al. 2006, 2008; Lemieux2008; Doussard et al. 2009; Heathcote et al. 2010; Cassiers & Keseloot 2012) and Canada(Green & Kesselman 2006; Frenette et al. 2007; Osberg 2008; Brzozwski et al. 2010; Bolton &3Breau 2012). This increase in inequality is evident at the urban scale with inequality risingbetween and within cities (Hamnett 2003: Volscho & Fullerton 2005; Korpi 2008; Borel-Saladin& Crankshaw 2009; Doussard et al. 2009; Van der Waal & Burgers 2009; Van der Waal 2010:Timberlake et al. 2012). Rising inequality leads to segregation of neighbourhoods based uponincomes, which can exacerbate the consequences of income inequality (Hulchanski et al. 2007;Watson 2009; Beeson et al. 2010; Reardon & Bischoff 2011; Chen et al. 2012; Ley & Lynch2012). Fundamentally, this thesis is a study of neighbourhoods, and the expression of growingincome inequality at the neighbourhood scale is of central concern.Chapter 3 is a quantitative study of regional income inequality at the first scale of analysis, theVancouver CMA. Three series of paired multivariate regressions are conducted using data fromthe 1996 and 2006 censuses in order to gain insight into significant associations betweendemographic characteristics and features of the housing stock, and the average income of CTs aswell as changes in average income.Chapter 4 focuses on the second scale of analysis, the emerging corridor of low and very-lowaverage income CTs that follow the SkyTrain Expo Line from East Vancouver through to NorthSurrey. The methodology of this chapter is a combination of semi-structured interviews with 14key informants (elected government officials, municipal and regional planners, and communitydevelopment and immigrant services professionals), supplemented by a review of newspaperaccounts and policy documents. All of the key informants have a professional mandate whichconcerns them with some part of the low income corridor. Several of the key findings of Chapter3 are supported by the dominant themes discussed by key informants, indicating that thecombination of quantitative and qualitative methods in this thesis is complementary.4As neighbourhoods are central to both Hulchanski et al. (2007) and Ley & Lynch (2012) andincreasing income inequality is expressed at the geography of the neighbourhood, certainneighbourhoods were selected for further study. Chapter 5 narrows the scale of analysis to twoneighbourhoods within the low income corridor. The dominant method of this chapter is a seriesof four focus groups with a total of 26 participants recruited by two community servicesorganizations operating in the corridor.Chapter 6 integrates the findings of chapters 3 through 5. The analysis of income inequality atthree scales in the Vancouver CMA, through the complimentary use of three distinct methodsrevealed common themes which are relevant at every scale of analysis.5Chapter 2: Literature ReviewThere is a general consensus that urban income inequality in developed nations has increasedsince the 1970s, but scholars’ conclusions regarding the causes of growing inequality varydepending on their theoretical framework and scale of analysis. This literature review considersthe theoretical frameworks of skills-biased technological change (SBTC) and the world andglobal cities hypotheses as explanations of increasing income inequality, and reviews a range ofempirical studies on urban income inequality in a variety of contexts.2.1 Theoretical FrameworksTwo broad explanations for income inequality emerged in the 1990s. Among economists, theconsensus of the early 1990s was that income inequality had grown during the 1980s and that theprimary cause of this was skill-biased technological change (SBTC) driven by the computerrevolution (Bound & Johnson 1992, Berman et al. 1994, Katz & Murphy 1992, Krueger 1993,Levy & Murnane 1992, Juhn et al. 1993). The SBTC thesis focused on differentials in humancapital and relative demand for skills while it rejected globalization as the main source of risinginequality.Another view which places globalization at the heart of rising inequality, particularly in cities, isthe ‘world’ or ‘global city’ thesis. First put forward by John Friedman (1986) and furtherdeveloped by Saskia Sassen (1988), the global city thesis predicts that the integration of the localurban economy into international flows of capital and people will lead to a growing share of bothlow-income and high-income occupations, with stagnant growth or decline for middle-incomejobs. This trend is otherwise known as the ‘polarization thesis.’ For both Friedman and Sassen,6the degree to which a city is central to the global economy will structure the class polarizationwithin. Economic restructuring , and high rates of international migration drive the gap betweenthe very high incomes of those who sit atop the urban occupational structure in professional andmanagerial services and the routine, low-wage, service-provision jobs often filled by women andimmigrants (see Friedman 1986 and Sassen 1988, 2001, 2006, 2008).Within the global cities literature, the polarization thesis has been problematized by a‘professionalization thesis,’ introduced by Chris Hamnet (1994, 1996). Hamnett argues that thepolarization thesis is conceptually underdeveloped, and fails to take into account the political andsocial context within which a given city exists. He argues that exposure to the global economy insome cities has led to the upgrading of skills in the local labour market, or that theprofessionalization of workers has occurred (Hamnett, 1996). The empirical basis of theprofessionalization thesis is covered in more detail below.2.2 Income Inequality at the International and National ScalesMotivated by dissatisfaction with available data on international income and wealthconcentration, Piketty & Saez (2006) constructed a database from historical tax statistics for anumber of mostly Western countries. Their data runs from 1917 to 2002, and they discover aremarkably similar trend for the US, UK, and Canada. During the inter-war period, the top tenpercent or decile of earners had a share of national income which fluctuated at around 40 to 45percent, but then declined sharply after the Great Depression and WWII. The top decile’s sharestayed relatively constant at 31 to 32 percent until the 1970s when it slowly began to rise. As of2002 the share of national income that went to the top ten percent of earners had recovered toclose to the pre-war level. Two of their other findings are especially of note: first, fluctuations in7income for the top ten percent can be accounted for mostly by fluctuations in income for the topone percent; second, the primary source of income for the very top earners before WWII wascapital income, whereas after the 1970s the source of income for the very top of earners hadshifted to salary income. At the national and international scale, growing income inequality islargely a result of growth in salary income for the top one percent of earners.Rising income inequality has not been a steady process since the 1970s. A number of studies(Autor et al. 2006, 2008; Lemieux 2008; Doussard et al. 2009; Heathcote et al. 2010) found thatin the 1980s, US income inequality grew at a faster rate than in the 1970s. Whereas the 1980ssaw inequality increase across the earnings distribution, in the 1990s inequality in the bottomhalf of the distribution stagnated and all recent inequality growth has occurred in the top half.Lemieux (2008) and Cassiers & Keseloot (2012) agree that the growth of the financial sector inthe economy can be implicated as a cause of this trend, while Autor et al. (2006, 2008) argue thatrapid growth in employment at the bottom and top end of the skills distribution relative to themiddle was due to the computerization of middle-income, routine cognitive tasks such asbookkeeping and repetitive production work, lending support to the SBTC thesis.But a political thesis could profitably be accommodated here as well. The 1980s are the decadein which the Regan-Thatcher leadership in the USA and the UK was association with strikinginstitutional reforms, ushering in the ideology and practice of neoliberalism. There is not spacehere to discuss this development in detail (see Harvey 2005; Peck and Tickell 2002), but sufficeto say it involved sweeping deregulation, the erosion of the welfare state, and the liberation ofmarket processes as the normative logic of policy decision-making. Changes to tax regimes wereadvantageous to the top 10 percent, while ‘welfare reform’ and later, after 2008, austerity8policies, have disproportionately penalised poorer groups (and poorer cities), reducing take-homeincome (Peck 2012).Frenette et al. (2007) and Brzozwski et al. (2010) show that income inequality in Canadaincreased substantially from the 1980s, and that much of the polarization in Canada has occurredwithin the extreme bottom and top of the income distribution. They use census data augmentedwith tax estimates to show that in Canada, inequality in after-tax disposable income issubstantially higher than previously thought. They cite institutional changes made in the 1990s totax rates and transfer payments (a key moment of federal neoliberalism) as being a leading causeof disposable income inequality. Due mostly to social assistance, income transfers, and childbenefit programs, disposable income inequality in Canada was basically flat until 1990, but sincethat time income inequality in Canada has been rising rapidly as federal policies attacked thenational deficit through welfare cuts and entrepreneurial initiatives (Bolton & Breau 2012; Green& Kesselman 2006; Osberg 2008). Canada is a multicultural country with a large visibleminority population and high rates of immigration. It has been found that Canadian-born visibleminorities face “substantial and significant earnings penalties” which have not diminished since1985, despite the considerable growth of this population (Hou & Coulombe 2010; Pendakur &Pendakur 2011).2.3 Income Inequality in CitiesEmpirical studies of income inequality in and between cities employ a variety of methods andoffer differing explanations for rising inequality. Glaeser et al. (2009) find that the dispersion ofeducational achievement is positively associated with income inequality and that in US citieshalf and one-third of the variance in income inequality could be explained by occupation-based9inequality and education respectively. Also consistent with the SBTC thesis was the finding thatcities with large numbers of both college graduates and high school drop-outs were especiallyunequal.Timberlake et al. (2012) set out to test the polarization thesis through a systematic analysis of 57large US cities. They found mixed support for the polarization thesis; income polarization wasnot clearly associated with centrality in the global economy, but when centrality was combinedwith high rates of immigration a positive relationship was discovered, which supports Sassen’s(2008) argument that low-paid service work at the bottom of the occupational hierarchy tends tobe filled by immigrants. In a study of 22 Dutch cities, Van der Waal (2010) found support for thepolarization thesis in that cities with a higher share of advanced producer services (a measure ofintegration into the global economy) had lower unemployment rates and a platykurtic incomedistribution.Deindustrialization and a decline in union density (another facet of neoliberalism) are oftenimplicated as a cause of growing income inequality in cities. Doussard et al. (2009) adapted amethod developed by Wright & Dwyer (2003) which codes industry-occupation groups and sortsthese groups into equally sized, wage- sorted quintiles. They examined changes in theoccupational structure of a number of US cities with a particular focus on Chicago. They wereable to show that polarization of Chicago’s labour market had occurred, and citeddeindustrialization and diminishing union density as a major cause of income inequality.Volscho & Fullerton (2005) found that income inequality was lower in cities with greater uniondensity and greater government sector employment. The positive benefit of union density wasstrongest for workers in the middle of the income distribution, and less so for those at the10bottom. They also found that unemployment was positively associated with income inequality, aswas greater dispersion in education and age. Further support for this explanation of growingincome inequality was found by Moller et al. (2009) who used US counties as their unit ofanalysis to show that rising inequality in the US reflects “the growth of large, deindustrializedurban areas.” Some of the most important factors influencing inequality in their study were thehigh school completion rate, the size of the black population, the size of the government sector,union density, and women’s labour force participation.The polarization thesis predicts that global cities will see growth at the top and bottom of theearnings distribution, with stagnation or shrinkage in the middle. Hamnett’s (2003) case study ofLondon, England suggests that growth in the relative size of groups at the top of the incomedistribution significantly outweighs growth at the bottom. Hamnett’s (2003) analysis has beencritiqued by Watt (2008) for relying upon occupational categories as a proxy of class, therebyexcluding the economically inactive. To this charge Hamnett (2009) responded that those left outof the analysis were largely excluded from the labour market because they were students, retired,or looking after the home and family, and that by no means should this group be assumed to beworking class. Davidson & Wyly (2012) counter the assertion that London could becharacterized as increasingly middle class and argue that conflating changing occupationalstructures with class relations is a mistake repeatedly made by Hamnett & Butler (2008). This isimportant because Hamnett’s claim that London’s working class population has been replaced byan expanding middle class depends upon the decline of traditional working class occupations.Davidson & Wyly (2012) argue that labelling the diverse occupations contained within socio-economic group (SEG) categories as broadly middle class is highly problematic, and the11application and interpretation of UK social class categories must be conducted in a more criticalmanner. They argue that Butler et al. (2008) lack an appreciation of the relational nature of class.In response, Hamnett and Butler (2013) argue that there is no viable alternative to usingcomprehensive official data sets because they can be used to consistently analyse change overtime and space.  In defence of their use of SEGs to demonstrate the growth of the middle class,they argue that the growth of professional and managerial groups was far more marked in InnerLondon than in either Outer London or the UK as a whole (Hamnett and Butler 2013). A keyinsight is that the debate over the professionalization thesis has largely been concerned withquestions of method, which shows how important decisions regarding data sources andmethodology are in forming conclusions in empirical studies of urban income inequality.Borel-Saladin & Crankshaw (2009) found support for the professionalization thesis through theirstudy of Cape Town, South Africa. Although Cape Town experienced an absolute growth inhighly paid professional jobs and an absolute increase in unskilled, low-wage occupations from1980 to 2000, for every job that was lost in the manufacturing sector, three middle-incomewhite-collar jobs were created. They argue that these white-collar middle-income jobs werecomparable in pay to the lost manufacturing jobs, and that other studies’ assumption that workersin the service sector make less than those in manufacturing may have led to an overestimation ofthe degree of polarization and an underestimation of the extent of professionalization.Further support for professionalization was found by Van der Waal & Burgers (2009) whofocused their analysis on firms located in Amsterdam and Rotterdam in the Netherlands. Byfocusing on individual firms they argued that they were able to form a direct connection betweenfirms’ levels of international integration and income inequality. They conclude that second-tier12cities such as Rotterdam experience professionalization because they are forced to compete moreaggressively in the international market.A unique explanation of urban income inequality comes from Korpi (2008), who found that thesize of a local urban labour market has a significant positive effect on the extent of incomeinequality, and that top level incomes were the most affected. These findings support a rank-sizerule thesis of income inequality, because labour markets become more diversified as populationincreases. However, Korpi admits that this may just be an observation of a spurious correlationrather than a causal insight.2.4 Income Inequality at the Neighbourhood ScaleIt has been observed that the residential segregation of people according to income, a processlabeled by Hulchanski et al. (2007) as ‘socio-spatial polarization’, is a consequence of incomepolarization. Watson (2009) argues that segregation of people along the lines of income hasincreased in the US since the 1970s. The 1980s saw the highest levels of income segregationbecause the poor faced “increasing concentration and isolation within the central city” as thewealthy segregated themselves into suburban neighbourhoods. As incomes become polarizedand concentrated in particular places, there is greater inequality between neighbourhoods(Reardon & Bischoff 2011). There is some evidence for path-dependency of inequality, as thosecities that were more unequal in 1980 tended to be more unequal in 2000 (Beeson et al. 2010).In order to map neighbourhood income segregation, Chen et al. (2012) adopted a methoddeveloped by Franette et al. (2007) to determine after-tax income for households and found thatneighbourhood income segregation in Canada’s eight largest cities had increased between 198113and 2006. In a study focused on the City of Toronto, Hulchanski et al. (2007) documentprocesses of socio-spatial polarization from 1970 to 2005. They argue that Toronto has becomedivided into three cities along lines of income and cite “changes in the economy, in the nature ofemployment (more part-time and temporary jobs), and in government taxes and incometransfers,” as the primary causes of “a growing gap in income and wealth and greaterpolarization among Toronto’s neighbourhoods.” Adopting Hulchanski et al.’s method, Ley &Lynch (2012) show similar processes of socio-spatial polarization in Vancouver neighbourhoods.It is worth noting that both Hulchanski et al. (2007) and Ley & Lynch (2012) find that the innercities of Toronto and Vancouver have experienced relative increases in income while somesuburban areas have experienced income decline. Oreopoulos (2008) found that manyhouseholds living in low-income areas are recent immigrants who tend to move out of such areaswithin five years. Given that in 2001, 62% of residents in areas of Toronto that experiencedrelative income decline since the 1970s were not born in Canada, and that in Vancouver lowerthan average incomes and income loss over time were associated with a high rate of non-Englishas mother tongue, immigration must play a role in Canadian neighbourhood socio-spatialpolarization.Oreopoulos (2009) conducted a field experiment with six thousand constructed resumes whichwere sent in response to online job postings across multiple occupations in Toronto. He foundthat interview requests rates were three times higher for English-named applicants with Canadianeducation and experience, than for resumes with non-English names with foreign education andexperience. However, he found that foreign applicants from Britain faced no such discrimination.All else held constant, Canadians with English-sounding names received interview requests 40%14more often than applicants with non-English-sounding names (11% vs. 8% respectively), whichmay be one reason that skilled immigrants struggle in Canada’s labour market.Whereas the empirical studies discussed above focus on the labour market and demographics toexplain neighbourhood income segregation, Glaeser et al. (2008) argue that having access topublic transportation is a key variable in explaining the concentration of the poor in particularneighbourhoods.The major findings of this thesis are generally supportive of the polarization thesis in globalcities. Consistent with the SBTC thesis, there has been a transition in the relative demand for jobskills as the importance of the industrial sector has declined in the City of Vancouver and therehas been a considerable rise in the importance of the service sector (Miro 2011). However, theVancouver CMA is one of three major landing sites of immigrants to Canada, and this studyfinds that polarization in the region is strongly associated with visible minorities and recentimmigration, which is far more consistent with the global cities’ argument that polarization isstrongly related to globalization and the immigration of visible minorities.2.5 Methods and DataIn the empirical studies above there is a wide variety of measures of inequality and a broad rangeof methods used to analyze data from various sources. Of the material surveyed in this literaturereview the most common sources of data were the US Census and the 5% or 1%  integratedpublic use micro-sample (Volscho & Fullerton 2005: Autor et al. 2006, 2008; Glaeser et al.2008, 2009; Lemiuex 2008; Beeson et al. 2010; Heathcote et al. 2010; Timberlake et al. 2012),and micro-data from the 20% sample long-form Canadian Census (Frenette 2007; Hulchanski et15al. 2007; Pendakur & Pendakur 2011; Chen et al. 2012; Ley & Lynch 2012). Some relied onsurvey data (Brzozoski 2010; Heathcote et al. 2010) while others created their own database(Korpi 2008; van der Waal 2010). The most commonly used measure of inequality was the Ginicoefficient (Glaeser et al. 2009; Moller et al. 2009; Reardon & Bischoff 2011; Chen et al. 2012)but many created their own measure (Hulchanski et al. 2007; Doussard et al. 2009; Watson 2009:Ley & Lynch 2012). In short, very few projects follow the same methodology, and I argue that inorder to study income inequality across contexts and time periods, methods must be consistent.For this reason I have chosen to build upon the methodology and findings of Hulchanski et al.(2007) and Ley & Lynch (2012). These studies provide a frame for researching incomeinequality in cities as a temporal process, and highlight neighbourhoods that are changing and inneed of further study.16Chapter 3: Quantitative Study of Income Inequality in the Vancouver CMAUrban income polarization is a complex and multi-scalar process. In order to account for thiscomplexity, this thesis is divided into three scales of analysis. The scale of analysis for thischapter is that of the region, using census tracts (CTs) as the geographical unit of measurement.The intention is to present a general context that forms the backdrop for the more narrowlyfocused chapters that follow. In this chapter extensive methods are used to interrogate incomepolarization in the Vancouver census metropolitan area (CMA).Three multivariate regression methods were chosen in order to gain insight into processes andconsequences of income polarization in the Vancouver CMA. In building the OLS model it isassumed that many of independent variables have a causal relationship with the dependentvariable, the average income of CTs. In the logistic regression models that follow, a causalrelationship between the independent and dependent variables is not assumed, but insight isgained into the characteristics and changes we could expect to find in CTs in which averageincomes have declined between 1971 and 2006.The majority of the selected independent variables are statistically significant in at least two ofthe regression models. The results of each regression are briefly discussed and a conclusionprovides an analysis of the entire series of six regressions, identifying patterns andcontradictions. The key finding of this chapter is that visible minority status is the mostimportant variable by far, suggesting that this is a key variable for understanding incomedisparity and income decline in Vancouver CMA. In conjunction with this finding,concentrations of recent immigration were strongly associated with income decline.17Through the selective testing of 15 independent variables in a series of six paired multivariateregressions, insight is gained into some of the factors which are associated with processes ofsocio-spatial polarization and the geographical distribution of low incomes in the VancouverCMA. The first series of regressions tests variables for statistical significance in relationship toaverage individual and household incomes at the CT level in 2006. In the second and third seriesof regressions the focus shifts to CTs which were identified by Ley & Lynch (2012) as bothexperiencing a 15% or greater decline in average individual or household incomes relative to theregional average, and being of low or very low income status in 2006. Income polarization iseffectively mapped and described by Ley & Lynch (2012) and this chapter aims to take theiranalysis one step further by highlighting associations between declining average incomes anddifferences between neighbourhoods.Three distinct methods are employed in order to identify statistical significance in theindependent variables using the SAS statistical software package:1. Ordinary least squares (OLS) linear regressions using 2006 census data.2. Logistic regressions using 2006 census data.3. Logistic regressions incorporating changes in the independent variables from 1996 to2006.3.1 Selection of Independent VariablesThe independent variables listed below were created using data from the 2006 Canadian LongForm Census. Data from the 2011 National Household Survey was not used due to a lack ofreliability for income data at the census tract scale (Hulchanski et al. 2013). The variables were18selected based on the expectation that they would be associated with the presence or absence oflow incomes. Recent publications on income dynamics include relevant findings: Walks (2014) found that in Vancouver individuals aged 65 and above were associatedwith lower debt-to-income-ratios, and the greatest decrease in the incidence of lowincome since 1980 has been amongst seniors (HRSDC 2014). Haan (2010) found that residential crowding in Vancouver is a housing affordabilitystrategy used most often by recent immigrants and visible minorities. Walks & Bourne(2006) found a connection between apartment housing, high levels of racial diversity,and the neighbourhood patterning of low income in Canadian cities. The link between recent immigration, visible minorities, and a prevalence of low incomein Vancouver has been established in Smith & Ley (2008). Lone parents are more at risk of experiencing low income that other Canadians (HRSDC2014). The development of new dwellings was included as a proxy for new-build gentrificationwhich transforms the characteristics of class in given areas (Davidson & Lees 2010). Glaeser et al. (2008) found a link between low incomes in cities and access to publictransit. Other independent variables were selected based on their association with the conditionof the housing stock, and the residential mobility of residents.3.1.1 The Independent Variables for 20061. Percentage of the population aged 65 and older.2. Percentage of households composed of 6 or more members.3. Percentage of the housing stock in apartment units in buildings five storeys or taller.4. Percentage of the housing stock in apartment units in buildings four storeys or less.5. Percentage dwellings that are rented.6. Percentage of the population that self-identifies as a visible minority.197. Percentage of households that are headed by a lone parent1.8. Percentage of the population that are recent immigrants, having landed between 2001 and2006.9. The unemployment rate.10. Percentage of individual income that is derived from government transfer payments.11. Percentage of dwellings in need of major repairs.12. Percentage of dwellings built from 1996-2000.13. Percentage of dwellings built from 2001-2006.14. Percentage of the working population that uses public transit to get to work.15. Percentage of the population that moved within 5 years.These variables were regressed in turn against the dependent variables, individual and householdaverage incomes for CTs reported for 2005. As I am using CTs as the scale of analysis thefollowing models do not provide insight into the incomes of individual people, but they do allowus to see what variables are shaping landscapes of wealth and poverty. Therefore, we gain insightinto the channeling factors that get people of a particular socio-economic status to be in one partof the city vs. another.3.2 Regression Series 1: OLS Multivariate RegressionsThe first series of regressions performed are simple, linear, OLS regressions. Of the 15independent variables listed above it was necessary to remove variable 14 (the percentage of theworking population that uses public transit to get to work) and variable 15 (percentage of thepopulation that moved within 5 years) in order to avoid problems of multi-collinearity. Tolerancestatistics have been included in table 1 as a measure of multi-collinearity. A tolerance statisticshows the variance in a given independent variable that is not predicted by variance in the other1 It should be noted that lone-parent families in Metro Vancouver in 2006 were overwhelmingly headed by women.20independent variables. In effect, each independent variable is regressed against the others. Atolerance statistic of less than 0.20 is considered to be too low to include in a regression model,as this indicates that more than 80% of the variance of a given independent variable can bepredicted by the variance in the other independent variables. In this model, the percentage ofdwellings that are rental has a tolerance statistic of 0.20. It was decided that rental housing wasan important variable and would be included despite strong multi-collinearity with the otherindependent variables.Independent variables 1 through 13 were tested for statistical significance and for strength ofrelationship to the dependent variable. The strength of this relationship is measured through astandardized estimate which indicates the effect that variance in the independent variable has onvariance in the dependent variable, ceteris paribus. A standardized estimate of 0.30 indicates thatan increase of one standard deviation in the independent variable will lead to an increase of 0.30of one standard deviation in the dependent variable. A standardized estimate of -0.30 indicatesthat a standard deviation increase in the independent variable will cause a 0.30 standarddeviation decrease in the dependent variable. In interpreting the standardized estimates of theindependent variables in this and all following regression models, the reader is reminded thatbeta values such as the standardized estimates show the specific impact of changes in anindependent variable when all other independent variables are statistically controlled. Forexample, the standardized estimate for visible minorities shows us the effect on the averageincome of a CT if there was an increase in visible minority persons whose aggregatecharacteristics would have no effect on the other independent variables in a CT.213.2.1 Dependent Variables of the OLS Regressions of Average Individual and HouseholdIncomes, 2006The dependent variable for the OLS regression of average individual incomes was created byimporting the average individual incomes of all 409 CTs of Vancouver CMA in the 2006 censusinto the SAS statistical analysis software package. The dependent variable for the OLSregression of average household income was created by importing the average householdincomes of the 409 CTs into SAS.3.2.2 Results of OLS Multivariate RegressionsTable 1 Results of OLS Multivariate Regressions3.2.3 Discussion of OLS Multivariate Regression ResultsThe OLS regression models do a reasonably good job of predicting variance in averageindividual and household incomes. The adjusted R² of 0.40 and 0.41 indicates that the modelAverage Individual Income Average Household IncomeAdjusted R-Square      0.40 Adjusted R-Square     0.41Variable Tolerance Standardized Significance Standardized SignificanceStatistic Estimate Level Estimate LevelIntercept **** ****Population aged 65+ (%) 0.57 0.29579 **** 0.23427 ****Households with 6 or more members (%) 0.37 -0.12266 -0.02963Housing units that are aparments in buildings 5+ storeys (%) 0.33 -0.11549 -0.21018 ***Housing units that are aparments in buildings 4 storeys or less (%) 0.30 -0.09549 -0.21031 ***Housing units that are rental (%) 0.20 -0.06392 -0.16204Population that self-identifies as a visible minority (%) 0.32 -0.25222 **** -0.14497 *Households headed by a lone-parent (%) 0.56 -0.28468 **** -0.26671 ****Population that is a recent immigrant (landing 2001-2006) (%) 0.39 0.04258 0.06413Unemployment rate 0.54 -0.01235 0.01172Individual income derived from government transfer payments (%) 0.61 -0.31949 **** -0.26572 ****Housing units in need of major repairs (%) 0.58 -0.02275 -0.01195Housing units that were built 1996-2000 (%) 0.67 -0.03910 -0.04873Housing units that were built 2001-2006 (%) 0.71 0.01904 -0.00687Number of observations 409 409*Statistically significant at the 5%; ** at 1%; *** at 0.1%; **** at 0.01%.Data Source: The Canadian Long Form Census, 2006.22successfully predicted 40% and 41% of the variance in individual and household averageincomes respectively.For both of the OLS regression models, the only statistically significant variable that had apositive effect on average incomes was the share of the population aged 65 and over. Alsoconsistent between the two regressions was the statistically significant and negative effect ofvisible minorities, lone-parent families and government transfer payments on average incomes.Apartments, whether in buildings 5 storeys and above, or in buildings 4 storeys and below, has astatistically significant, negative association with average household incomes. As this was notthe case for individual incomes, this result suggests a link between housing type and householdincomes. It is possible that this could be the result of the lower housing cost of apartments, ordue to the presence of more single-earner households in apartments.It should also be noted that the coefficient for visible minorities was much stronger for individualincome than for household income. This is entirely consistent with Ley’s (1999) findings of thedifferences between the rank of individual and household incomes for ethnic minority groups inthe Vancouver CMA. There are multiple earners in some visible minority households thatdiminish the effect of individual low incomes. A key consideration here that relates to thefindings for both apartments and visible minorities is that household incomes, not individualincomes, determine where people can afford to live.It is worth noting that the percentage of dwellings that are rental is not statistically significant inthis model. This is likely a result of suppression due to the high rate of multi-collinearity with theother independent variables. This is particularly true for apartments in buildings 5 storeys and23taller and for apartments in buildings 4 storeys or less, which have a correlation coefficient withrental dwellings of 0.55 and 0.67 respectively. That rental housing has a negative relationshipwith the average income of CTs is supported by the correlation coefficients between thepercentage of dwellings that are rental and the average individual and household income of CTsat -0.32 and -0.52 respectively.The weak coefficient against new immigrants is unexpected, but is suppressed due to collinearitywith the statistically significant visible minority variable (r=0.63). The unemployment ratevariable is weaker than expected as it has a strong correlation coefficients with the statisticallysignificant variables of visible minorities (r=0.40) and lone-parent families (r=0.49).3.3 Regression Series 2: Logistic Regressions Using 2006 Census DataThe method developed in Hulchanski et al. (2007) and employed in Ley & Lynch (2012)compares changes in the average incomes of CTs to changes in the regional average between twocensus years; that of 1971 and 2006. In order to allow for comparison between the census years,the CT boundaries of 2006 and the data within these boundaries were altered so that theboundaries and data conformed to those of 1971. In the case of the Vancouver CMA, those CTsin which average incomes had increased or decreased by 15% or more relative to the regionalaverage income between 1971 and 2006 were considered to be significant, and mapped. Becausethis method treats income as relational, it automatically controls for inflation.24Figure 1 Change in Census Tract Average Individual Income, 1970-2005. Source: Ley & Lynch (2012)(reproduced with permission)Figure 2 Change in Census Tract Average Household Income, 1970-2005. Source: Ley & Lynch (2012)(reproduced with permission)25Much of this thesis is guided by the findings of Ley and Lynch (2012) and as a furtherinterrogation of these findings I sought to discover correlations between the independentvariables and the CTs in which the average income had declined between 1971 and 2006. CTs inwhich average incomes had increased relative to the regional average are indicated in blue infigures 1 & 2, and those CTs in which average incomes had decreased are marked in brown. Adistinction was made between two types of CTs in which incomes had declined; those that werein low or very income status in 2006 are indicated in solid brown while those in middle or higherincome status in 2006 are indicated in hatch-marked red/brown. A low income CT was definedas one with an average income of 80% to 60% of the regional average, and a very low incomeCT was defined as having 60% or less of the regional average income in 2006. The focus of thelogistic regressions below are the CTs which had experienced relative average income decline of15% or more from 1971 to 2006 and were in low or very low income status in 2006; in otherwords, all of the CTs marked solid brown in Ley & Lynch (2012). For the sake of brevity, allCTs in which incomes declined by 15% or more relative to the regional average and were in lowor very low income status in 2006 will be hereafter referred to as ‘experiencing individual orhousehold income decline.’It was beyond the scope of this research project to convert 2006 census data to conform to 1971census data, but it was necessary to establish a method that would make it possible to relate thefindings of socio-spatial polarization in Ley & Lynch (2012) to the boundaries and data of the2006 census. Over the course of 35 years, the 178 CTs of Vancouver CMA in 1971 had beendivided into 409 CTs by 2006. The splitting of CTs is a practice of Statistics Canada to keep thepopulation of a CT around the average of 5000 individuals. As the population of Vancouver26CMA grew significantly from 1971 to 2006, the result is a large increase in the number of CTs.Despite the remarkable increase in the number of CTs during this period, the 409 CTs of 2006can be easily related back to the CT boundaries of 1971, as new CTs normally remain within theboundary of the CT from which they are split. Being able to relate the CTs of 2006 to the CTs of1971 made it possible to employ a logistic regression methodology to bridge the gap between thefindings of Ley and Lynch (2012) and data in the 2006 census.A logistic regression is a multivariate regression in which the dependent variable is binary; inthis case either a ‘1’ to mark an event or a ‘0’ to mark a non-event. Those CTs of 2006 whichwere within the 1971 CT boundaries that experienced income decline were coded as an event,and all other CTs were coded as a non-event. This will be explained further for each of thelogistic regressions. The independent variables listed above and included in the OLS regressionmodels were used in the following logistic regressions. Variables 14 and 15 were excluded fromthese models to avoid problems of multi-collinearity.3.3.1 Dependent Variable for Logistic Regression of Individual Income DeclineIn the map above of average individual income change in Ley & Lynch (2012), 23 of the 178CTs according to the boundaries of 1971 are identified as experiencing individual incomedecline. According to the CT boundaries of 2006, there are 73 CTs that fall within these 23 CTsof declining individual income. In order to create a dependent variable for this regression, these73 CTs in 2006 were coded with a ‘1.’ Hereafter these CTs will be referred to as being within the27boundaries of individual income decline. All other CTs were coded with a ‘0.’ Of the 409 CTs in2006, 73 were coded as an event and 336 were coded as a non-event2.3.3.2 Dependent Variable for Logistic Regression of Household Income DeclineIn order to create the dependent variable for average household income decline, the processdescribed above was repeated according to the map of household income change in Ley & Lynch(2012). There are 60 CTs in 2006 that fall within the 1971 CT boundaries of household incomedecline. Hereafter these CTs will be referred to as being within the boundaries of householdincome decline. These CTs were marked with a ‘1,’ and all others were marked with a ‘0.’ Thisprocess created a dependent variable with 60 events and 349 non-events.2I tested whether 2006 CTs that were of low- or very low-income status according to 1971 CT boundaries fit thatdefinition. For average individual incomes in 2006 the designation proved to be accurate with 60 of 73 CTs (86%)contained within the 1971low-income boundaries conforming to the definition of low income. For averagehousehold incomes 36 of 60 CTs (60%) within the 1971 boundaries met the definition. The neighbourhood effectsof particularly low household income areas were sufficient to bring down the average household income of CTswhen reformatted to 1971 boundaries.283.3.3 Results of Multivariate Logistic Regressions of Income DeclineTable 2 Results of Multivariate Logistic Regressions of Individual and Household Income Decline3.3.4 Discussion of Multivariate Logistic Regressions ResultsIn discussing these results it is important to note that the independent variables are drawn from2006 data and the dependent variable is a proxy of income change from 1971 to 2006. Due thetemporal nature of the dependent variable it is not possible to claim causality in these models,because the independent variables observed in 2006 could not have a causal relationship withprocesses of income decline beginning in 1971. What these models can show us are thecorrelations between the independent and dependent variables. It should also be noted that insomething of a reversal of normal statistical logic, the dependent variable is ‘explaining’ theindependent variables; that is to say if the average income of a CT declined, the independentIndividual Income Decline (1971-2006) Household Income Decline (1971-2006)Low to Very Low Income Status in 2006 Low to Very Low Income Status in 2006Max-rescaled R-Square 0.60 Max-rescaled R-Square 0.48Variable Standardized Significance Odds Standardized Significance OddsEstimate Level Ratio Estimate Level RatioIntercept **** ****Population aged 65+ (%) -0.1235 0.799 -0.3857 ** 0.497Households with 6 or more members (%) 0.6262 *** 3.113 0.6834 **** 3.454Housing units that are aparments in buildings 5+ storeys (%) 0.1608 1.339 0.8094 **** 4.341Housing units that are aparments in buildings 4 storeys or less (%) 0.2102 1.464 0.5325 ** 2.627Housing units that are rental (%) -0.2913 0.590 -0.5304 ** 0.382Population that self-identifies as a visible minority (%) 0.6134 ** 3.042 -0.1659 0.74Households headed by a lone-parent (%) 0.2358 1.534 0.2701 1.632Population that is a recent immigrant (landing 2001-2006) (%) 0.711 **** 3.632 0.602 *** 2.98Unemployment rate -0.0432 0.925 -0.0533 0.908Individual income derived from government transfer payments (%) 0.091 1.179 0.2771 * 1.653Housing units in need of major repairs (%) 0.3908 ** 2.032 0.0741 1.144Housing units that were built 1996-2000 (%) -0.0272 0.952 -0.1487 0.764Housing units that were built 2001-2006 (%) -0.2011 0.694 -0.1713 0.733Number of observations 409 409Number of events 73 60Number of non-events 336 349*Statistically significant at the 5%; ** at 1%; *** at 0.1%; **** at 0.01%.Data Source: The Canadian Long Form Census, 2006.29variables tell us what we could expect to see there, but not what caused average incomes todecline.In order to facilitate interpretation, an odds ratio has been provided. The odds ratio is a measureof the strength of the relationship between an independent and the dependent variable in alogistic regression. An increase of an independent variable in a given CT will affect the odds thatthis CT will be within the boundaries of income decline by a factor of the odds ratio to one,ceteris paribus. For example, an odds ratio of 2 indicates that a standard deviation increase of theindependent variable within a given CT will increase the odds that the CT will be within theboundaries of income decline by a factor of 2 to 1. Alternately, an odds ratio of 0.5 indicates thata standard deviation increase in the independent variable will decrease the odds of a CT beingwithin the boundaries of income decline by a factor of 0.5 to 1. An odds ratio of 1 indicates anextremely weak relationship.The logistic regression model for average individual income decline performed remarkably well,returning a max-rescaled R² of 0.60. The logistic regression of average household income declineperformed quite well, returning a max-rescaled R² of 0.48.The percentage of the population aged 65 or over had a negative influence on the odds. In a CTin which incomes had declined, we would expect the percentage of the population aged 65 andover to decline as well. This result is consistent with the OLS regressions above, in which therewas a statistically significant relationship between an aging population and higher average CTincomes.30For both individual and household incomes, households of 6 members or more and recentimmigrants were statistically significant with a strong positive odds ratio. For CTs in whichincomes declined we could expect to see an increase in these variables.In the regression of average individual income decline, visible minorities and housing in need ofmajor repair were statistically significant and increased the odds ratio. In CTs in whichindividual incomes had declined we could expect to see an increase in the share of visibleminorities and housing in poor condition.In the logistic regression of household income decline, housing type proved to be of muchgreater importance. The largest odds ratio in either of the logistic regressions above was 4.34 forapartment units in buildings 5 storeys tall or above. Apartment units in buildings 4 storeys or lesswas also statistically significant, but with a weaker odds ratio. In CTs in which householdincomes had declined we could expect to see an increase in the share of apartments within itsboundary. This result is consistent with the results of the OLS regression in which apartmentunits had a statistically significant and negative influence on average household incomes.That visible minorities were not significant in the regression of average household incomedecline could be explained by a higher propensity for visible minorities to have largerhouseholds with multiple earners, which obscures the lower individual earnings of visibleminorities (Ley 1999, Haan 2010).It is somewhat puzzling that CTs in which household incomes had declined, we could expect tosee a lower share of dwellings that are rental. It was assumed that rental housing would be31associated with lower incomes, but as in the case of the previous OLS regressions, it is possiblethat this variable is being suppressed due to multi-collinearity with the apartment variables.3.4 Regression Series 3: Logistic Regressions of Change, 1996-2006The method developed in Hulchanski et al. (2007) and employed in Ley & Lynch (2012) is anexamination of changes in income over time. In keeping with their method it was deemednecessary to conduct a series of logistic regressions that could take into account correlationsbetween changes in the independent variables and CTs which experienced income decline.Examining how the independent variables had changed from 1971 to 2006 would offer littleinsight into recent processes of income polarization in Vancouver CMA, and as inequality inCanada began to increase sharply in the 1990s (Conference Board of Canada 2011) the decisionwas made to examine changes in the variables between 1996 and 2006. The independentvariables listed below are based upon the independent variables used in the previous OLS andlogistic regressions, but these variables have been transformed in order to reflect change overtime. As discussed earlier, in 2006 there were 409 CTs, and in 1996 there were 298. In order tomake the CT boundaries and data of 2006 conform with those of 1996 it was necessary to re-joinsome of the 2006 CTs so that they could be compared to the CTs from which they were splitafter 1996. By looking through Statistics Canada Census Conversion Files I was able todetermine which of the CTs in 2006 had been created in both 2001 and 2006. In order to re-join aCT, I created a weighted average of the independent variables (using relevant statistics asweight) and then converted the data into percentages. In this way I was able to reduce the ‘n’ of2006 CTs from 409 to 298. It was then a simple matter of subtracting the percentage values ofthe independent variables in 1996 from the percentage values of the independent variables in322006 to create independent variables of change from 1996 to 2006. An inevitable consequence ofre-joining CTs that were split because of population growth is that CTs with much largerpopulations were given an equal weight to CTs with much smaller populations3.When the independent variables were converted to reflect changes from 1996 to 2006, there wereno issues of multi-collinearity as there was little correlation in how the variables changed overtime. Therefore, variables 14 and 15 could be included in this series of logistic regressions.1. Change in the percentage of the population aged 65 and over.2. Change in the percentage of households composed of 6 members or more.3. Change in the percentage of housing units which are apartments in buildings 5 storeys tallor above.4. Change in the percentage of housing units which are apartments in buildings 4 storeys tallor less.5. Change in the percentage of housing units which are rental.6. Change in the percentage of the population which self-identifies as a visible minority.7. Change in the percentage of households which are lone-parent.8. Change in the percentage of the population which is recent immigrant, those havinglanded between 1991-1996 and 2001-2006.9. Change in the unemployment rate4.10. Change in the percentage of individual income derived from government transferpayments5.3 Some notable outliers in terms of re-joined 2006 CTs which conform to 1996 boundaries and have very largepopulations are as follows: 0059.02 (2006 pop. 19,665), 0147.03 (2006 pop. 17,455), 0182.00 (2006 pop. 17,220) 6CTs re-joined, 0183.02 (2006 pop. 19,970) 4 CTs re-joined, 0185.02 (2006 pop. 17,145), 185.03 (2006 pop.22,345), 187.02 (2006 pop. 27,450) 5 CTs re-joined,  0188.00 (2006 pop. 19,545) 6 CTs rejoined, 287.03(2006 pop.27,595) 6 CTs rejoined, 0400.00 (2006 pop. 17,490).4 When rejoining CTs in 2006 which were split from CTs in 1996, a weighted mean for the unemployment rate wascreated according to the total population aged 15 and above in the labour force.5 In re-joining 2006 CT data to conform to 1996 boundaries, a weighted mean of this variable was determined usingtotal individual income as the weight. One CT of special interest in terms of government transfer payments was CT0059.01. Between the 1996 and 2006 census this CT was split into 0059.05 and 0059.06. The percentage of3311. Change in the percentage of dwellings in need of major repairs.12. The percentage of housing units which were developed between 1996 and 2000.13. The percentage of housing units which were developed between 2001 and 2006.14. Change in the percentage of workers who use public transit to get to work.15. Change in the percentage of the population which had moved within 5 years.The dependent variables for this series of regressions were built in a similar way to the previousseries of logistic regressions. Those 1996 CTs which were within the 1971 CT boundaries thatexperienced individual and household income decline were coded with a ‘1’ and all other CTswere coded with a ‘0.’individual income derived from government transfers for 0059.01 in 1996 was 22.1%. In 2006 the percentage ofindividual income derived from government transfer payments in CT 0059.05 and 0059.06 was 4.2% and 38%respectively. This is an intriguing example of the modifiable areal unit problem in which geographical boundariescan have a significant effect on results.343.4.1 Results from Multivariate Logistic Regressions of Change, 1996-2006Table 3 Results from Multivariate Logistic Regressions of Change, 1996-2006The multivariate logistic regressions of change performed quite well, returning a max-rescaledR² of 0.41 and 0.42 for individual and household income decline respectively. The variable withthe strongest association to both individual and household income decline was visible minorities.In fact, the odds ratio of 4.591 for individual income decline is the strongest in any of the logisticregressions. Therefore, in CTs in which incomes had declined from 1971 to 2006 we couldexpect to see an increase in the concentration of visible minorities over the 1996-2006 period. Itis worth noting that in this series of logistic regressions, visible minority status is statisticallysignificant in relationship to household income decline. It appears that by measuring changes inthe concentration of visible minorities, the obscuring effect of residential crowding and multipleearners is overcome.Individual Income Decline (1971-2006) Household Income Decline (1971-2006)Low to Very Low Income Status in 2006 Low to Very Low Income Status in 2006Max-rescaled R-Square 0.41 Max-rescaled R-Square 0.42Variable Standardized Significance Odds Standardized Significance OddsEstimate Level Ratio Estimate Level RatioIntercept *** ****Change (%) population aged 65+ 0.0948 1.188 0.0881 1.173Change (%) households with 6 or more members -0.3432 * 0.537 -0.2166 0.675Change (%) housing units that are aparments in buildings 5 storeys tall or more -0.1449 0.769 -0.1635 0.743Change (%) housing units that are aparments in buildings 4 storeys tall or less -0.0701 0.881 -0.2550 0.630Change (%) housing units that are rental 0.2182 1.486 0.1399 1.289Change (%) self-identifies as a visible minority 0.8403 **** 4.591 0.6194 **** 3.075Change (%) households headed by a lone-parent -0.0618 0.894 -0.0040 0.993Change (%) recent immigrant (landing 1991-1996 and 2001-2006) -0.2818 * 0.600 0.1843 1.397Change in the unemployment rate -0.3088 ** 0.571 -0.3779 ** 0.504Change (%) individual income derived from government transfer payments -0.2485 0.637 -0.0969 0.839Change (%) housing units in need of major repairs 0.3402 * 1.853 0.1542 1.323New housing units that were built 1996-2000 (%) -0.0920 0.846 -0.0491 0.915New housing units that were built 2001-2006 (%) 0.0097 1.018 -0.2037 0.691Change (%) workers who use public transit to get to work -0.1065 0.824 -0.0459 0.920Poplation who had moved within 5 years (%) 0.3072 1.746 0.5629 ** 2.776Number of observations 298 298Number of events 46 36Number of non-events 252 262*Statistically significant at the 5%; ** at 1%; *** at 0.1%; **** at 0.01%.Data Source: The Canadian Long Form Census, 1996; 2006.35Surprisingly, for both individual and household incomes, CTs which experienced a decline inincome could be expected to have a shrinking unemployment rate between 1996 and 2006. Thiscan also be thought of in the reverse, that a CT with a declining unemployment rate had greaterodds of being within the boundary of declining incomes. Further investigation revealed that theunemployment rate in 1996 was 8.6, whereas the unemployment rate for the Vancouver CMA in2006 was 5.6 (Metro Vancouver 2013). The four CTs which experienced the greatest decline inthe unemployment rate were 0058.00 (-17.5) located in the DTES, 0059.01 (-17.0) in downtownjust west of the DTES, 0185.05 (-12.3) located in the Newton neighbourhood of Surrey, and0190.03 (-12.2) located in the Whalley neighbourhood of Surrey. Except for 0059.01, these CTsexperienced household income decline into low income status, and the two CTs in Surreyexperienced individual income decline into low income status. That the unemployment rateshould decline the most in neighbourhoods which are recognizable for their low-income statussuggests that this variable reflects changes in the labour market, increased barriers tounemployment benefits and the rise of the working poor (Mendelson & Battle 2011; Stapleton etal. 2012). When this result is considered in conjunction with visible minorities, we can see thatan increasing concentration of visible minorities and a decreasing unemployment rate have astrong association with income decline. This coincidence is consistent with the world and globalcities hypotheses in which racialized minority groups work for diminishing returns (Friedman1986; Sassen 2001).An increase in the percentage of dwellings in need of major repairs was statistically significant inrelationship to individual incomes, with a fairly strong odds ratio, which comes as little surprise36as areas that experience income decline would be expected to experience a progressive decline inthe quality of the housing stock.Households with 6 or more members and recent immigrants both have a statistically significant,negative association with individual income decline which is somewhat counterintuitive whenwe consider the previous series of logistic regressions. In the logistic regression of 2006 data, wewould expect to see more households of 6 or more and recent immigrants in a CT which hadexperienced both individual and household income decline. However, when we consider thatthese are change variables, meaning that an increase in either variable would reduce the odds ofan association with income decline, we can reconcile these findings with the previous logisticregressions. The percentage of households that had 6 members or more declined from 4.4% in1996 to 4.1% in 2006, and the percentage of individuals who were recent immigrants declinedfrom 10.5% in 1996 to 7.2% in 2006. There was an overall relative decrease in both of thesevariables and those CTs which saw the decrease in either variable between 1996 and 2006 mayhave already had a high percentage in 1996. For example, the two CTs which saw the greatestdecrease in the percentage of six-person plus households (0225.00 and 0187.01) also had thehighest percentage of six-person plus households in 1996 (22.2% and 17.1% respectively).Further, all of the five CTs which saw the greatest decrease in the percentage of recentimmigrants from 1996 to 2006 were in the top ten CTs for the percentage of recent immigrants in1996.Variables 14 (workers taking public transit to work) and 15 (percentage of the population thatmoved within 5 years) were excluded from the first two series of regressions due to problems ofmulti-collinearity. When included as a measure of change in this series of regressions, changes in37the percentage of workers who took public transit to get to work was not significant. An increasein the percentage of the population that had moved within the last five years was significant atthe 1% level, and modestly increased the odds. It is plausible that some CTs which haveexperienced household income decline are attractive to new residents with lower householdincomes.3.5 ConclusionThe majority of the independent variables were statistically significant in at least two of theregression models. The only variable that had a positive association with both individual andhousehold incomes was the percentage of the population that was aged 65 or older, and thisvariable reduced the odds that a CT would be within a boundary of household income decline.These results are consistent with the findings of Walks (2014) and the fact that the incidence oflow incomes in seniors has been decreasing in the last 30 years (HRSDC 2014).The story that emerges from the remaining results of statistical significance is consistent with theworld and global cities hypotheses (Friedman 1987; Sassen 2001). Recent immigrants, especiallyif they are of visible minority status, face an income disadvantage which is compounded bydifficulty accessing government transfer payments. As they have a lower propensity than theCanadian-born to draw from social services, many are forced into the ranks of the working poorin the tertiary sector. In this context it is necessary to engage in housing affordability strategies,such as residential crowding of multiple income generators in deteriorating dwellings, withapartments being of particular significance to households.38Although recent immigration was not statistically significant for average individual andhousehold incomes in 2006, we would expect to see more recent immigrants in CTs whichexperienced household income decline between 1971 and 2006. The percentage of thepopulation that self-identified as a visible minority was significant in 5 of the 6 regressionsperformed. Clearly, the residential distribution of visible minorities in the Vancouver CMA hasmuch to tell us about socio-spatial polarization. Hiebert et al. (2006) found that immigrants andvisible minorities, whether owners or tenants, are more vulnerable than the Canadian-born inVancouver’s housing market, and that visible minorities are overrepresented in the high riskcategory of devoting more than half of their income to housing. In addition Smith and Ley(2008) found that immigrant status (and therefore visible minority status) was becoming a moreimportant correlate of poverty in Vancouver and Toronto from 1991 to 2001.In the first series of OLS regressions both the percentage of individual income that was derivedfrom government transfer payments and the percentage of households that were headed by a loneparent had a negative effect on both the average individual and household incomes of CTs. Itmust be noted that the vast majority of lone-parent households in the Vancouver CMA in 2006were headed by women. CTs which experienced household income decline could be expected tohave a higher share of income derived from government transfer payments, but the strongcorrelation in 2006 between government transfer payments and the elderly (r=0.50) due topensions complicates this relationship.The unemployment rate was not significant in the OLS regressions or in logistic regression series2. However, in CTs which experienced individual or household income decline, we could expectto see a decrease in the unemployment rate from 1996 to 2006. There are several possible39explanations of these results; changes in the labour market, the increasing difficulty of qualifyingfor benefits such as employment insurance and welfare, the lower propensity of immigrants todraw upon social services than their Canadian-born counterparts, or the rise of the working poor(Hiebert 2006; Mendelson & Battle 2011; Stapleton et al. 2012).Households of 6 members or more did not have a significant effect on average incomes in 2006,but we could expect to see more households of 6 or more in CTs which experienced individual orhousehold income decline. Visible minorities are far more likely than the Canadian-born toexperience residential crowding in the Vancouver CMA, and they are more likely to live inhouseholds with multiple earners which explains why visible minorities were not significant inthe logistic regression of household income decline (Ley 1999; Haan 2010).In CTs in which individual incomes declined, we could expect to see more dwellings in need ofmajor repairs and an increase in the percentage of dwellings in need of major repairs but thisvariable was not significant in any of the regressions of household income decline.Consistent with the findings of Walks & Bourne (2006) apartments had a statistically significanteffect on household incomes. A prevalence of apartments was associated with lower averagehousehold incomes in 2006, and those CTs in which household incomes declined would beexpected to have a proportionally large share of apartments. This relationship may be due to thedemographic characteristics of apartment occupants, the relatively lower cost of apartments tohouses, or as apartments are typically smaller than other forms of housing, they may house fewerincome earners.40The complementary findings for visible minorities, recent immigration, government transferpayments, the unemployment rate, and residential crowding in deteriorating apartment units areall consistent with theories of income polarization within global cities. The finding of thischapter offers further support to the world and global cities’ argument that income polarization isa starkly racialized process (Friedman 1987; Sassen 2001).The development of new dwellings, whether completed during 1996-2000 or 2001-2006, was notsignificant in any of the regression models. This is a surprising result as it was expected that thepresence of recent development would contribute to polarization through new-buildgentrification (Davidson & Lees 2010). Future research should examine the relationship betweendevelopment and increasing average incomes to determine if these variables have an effect onincome polarization at the upper end of the income distribution.It was surprising that rental housing was not statistically significant in any of the regressionsexcept for the logistic regression of household income decline, and a further surprise was thatwithin CTs that had experienced income decline we would expect to see a reduction in thepercentage of dwellings that are rental. I had assumed that a higher percentage of rentaldwellings would lead to lower and declining incomes, but due to suppression and multi-collinearity the interpretation of rental housing in this model is problematic.This chapter must conclude with a measure of caution. As I adapted 1996 and 2006 census datato allow for their comparison with CTs experiencing declining average incomes from 1971-2006as identified by Ley & Lynch (2012), the dependent variables created for the four logisticregressions are merely a proxy of income decline, not a precise measure of income change overtime. A more rigorous approach would develop a dependent variable of percentile changes in41average incomes for a variety of time periods. In addition to this it would be necessary tocalculate changes in the independent variables so that, for example, percentage changes inaverage income and other independent variables between 1986 and 2006 could be compared tosimilar processes between 1996 and 2006. An undertaking such as this would require asignificant investment of time and resources, which sadly was beyond the scope of this currentresearch project. The development of this method could be of use in a variety of contexts as itcould be applied to many Canadian CMAs to generate valuable insights into processes of incomepolarization across Canada.42Chapter 4: The Emerging SkyTrain Low-Income CorridorIn the time period 1971-2006 there were significant changes in the prevalence and distribution oflow-income CTs in the Vancouver CMA. These changes are the focus of this chapter. In figure 3below, Ley & Lynch (2012) mapped average individual incomes in 1971; low income censustracts are defined as having an average individual income of 60-80% of the CMA average, andvery low income census tracts are defined as having an average income of 60% or less of theCMA average. As we can see, the majority of census tracts fall within the middle-incomecategory, there is a concentration of low and very low income census tracts in downtownVancouver, with a few outliers of low income in New Westminster, Coquitlam, Surrey, andRichmond.43Figure 3 Census Tract Average Individual Income, 1971. Source: Ley & Lynch (2012) (reproduced withpermission)Figure 4 below maps average individual incomes in 2006 and shows that the number of middle-income CTs had shrunk significantly relative to 1971. Along with the decrease of middle-incomeCTs there was an increase in the number of both high- and low-income CTs. Notice theemergence of a low income corridor that roughly follows the Skytrain Expo Line; starting in EastVancouver, stretching through South Burnaby, being briefly interrupted in New Westminster, butthen carrying into North Surrey.44Figure 4 Census Tract Average Individual Income, 2006. Source: Ley & Lynch (2012) (reproduced withpermission)The trend is repeated in figures 5 and 6 when average household incomes are measured. In 1971the majority of census tracts fall in the middle-income category, but by 2006 there is evidence ofpolarization and the presence of a low-income corridor is even more pronounced along theSkytrain Expo Line. It is this emerging low-income corridor which is the focus of this chapterand the second scale of analysis in this thesis.45Figure 5 Census Tract Average Household Income, 1971. Source: Ley & Lynch (2012) (reproduced withpermission)Figure 6 Cesus Tract Average Household Income, 2006. Source: Ley & Lynch (2012) (reproduced withpermission)46The primary source of data for what follows are semi-structured interviews conducted with 14key informants, all of whom have professional work mandates in a neighbourhood or severalneighbourhoods within the corridor. Key informants are elected government officials, regionaland municipal planning professionals, and community development and immigrant servicesprofessionals. The names of key informants have been coded to protect their identities. Theinterviews are supplemented with newspaper accounts and policy documents.The main finding of this chapter is that the low income corridor must be understood via thepresence of three key factors, their inter-relationships, and their interaction with the SkyTrainLine. The first factor in explaining the presence of low incomes in the corridor is the presence ofrental housing in a variety of forms. Of significance to particularly low incomes is the presenceof aging three- to four-storey, wood-frame, purpose-built rental housing (PBRH). The secondfactor, which is closely related to the first, is the occupancy of rental housing by vulnerablegroups, particularly recent immigrants and refugees. The third factor is the presence ofcontingent development near public transportation stations in every municipality that theSkyTrain Expo Line runs through. This development is contingent because it varies greatly in itsform, process, and community response, but throughout the corridor neighbourhood change istaking place; creating issues where development pressures come to bear upon areas ofdisinvestment, particularly when these are areas of affordable rental housing occupied byvulnerable groups. The interactions between these factors vary across the four municipalities thatthe corridor runs through and this is illustrated through a brief discussion of these interactions ineach municipality.474.1 Origins of the CorridorThe Kingsway and SkyTrain Expo Line corridors have their roots in BC’s colonial era. In 1860 atrail was cut through the forests of what would later become the Municipality of Burnaby. Itspurpose was to allow for the movement of European troops and settlers from New Westminsterto the salt waters of False Creek. In 1872 this path was widened to allow for the passage of ahorse team and renamed Vancouver Road. In 1891 the communities of New Westminster andVancouver were connected by a tramway. The orange line in figure 7 below indicates the path ofthe Inter-Urban Line that operated until 1954. Settlement along Vancouver Road was sparse until1913 when further improvements were made, and the thoroughfare was renamed Kingsway(Beasley 1976).Figure 7 Route of the Inter-Urban Line. Source: Wrigley Street Map (reproduced with permission from theBurnaby Village Museum)48A significant share of the Vancouver region’s settlement patterns and movements wereestablished by this ‘trail through the forest’ as it cuts diagonally across the grid system, making itan efficient route for rail, rapid transit, and automobile traffic. It is along this right of way fromMount Pleasant in Vancouver to the waterfront of New Westminster that the Skytrain Line wasrouted in the early 1980s. The degree to which these movement possibilities have influenced theemergence of a low income corridor is difficult to quantify, but a qualitative link was assumedby several key informants.At first blush when you look at that corridor, you associate it immediately with the railcorridor that already existed. It’s not surprising that lower incomes are around railcorridors, because rail corridors generally decrease property values. So you find that veryoften that kind of a corridor will have lower income around it (B4).I think one of the things that is really interesting to explore is that because the rapidtransit links these older historic centres where there has already been purpose built(rental) housing or there’s lower income people who have lived in those areas. Is itbecause transit already linked these areas together, or has it attracted more lower incomepeople who rent to those areas because of higher levels of accessibility by transit itaffords? (R2).Rather than being a cause of low incomes, the route of the Skytrain Line may be a result of bothan existing rail right-of-way and the existing presence of low incomes. The high-incomeneighbourhoods of Shaugnessy and Kerrisdale were able to successfully resist a rapid transit linethat was proposed to follow an existing rail right-of-way to connect Richmond and theVancouver International Airport to Downtown Vancouver for the 2010 Olympic Winter Games(Chrunik 2007). However, resistance to the SkyTrain Line in the Vancouver neighbourhood ofGrandview-Woodland, a traditionally lower-income area, was unsuccessful.49Typically one of the things that happens with rail lines and public transportation is theydon’t go through areas that have higher incomes. There’s kind of a self-fulfillingprophesy that happens there, that areas that have higher incomes are insulated [from rapidtransit infrastructure] (B4).There was quite a lot of mobilization around the Expo Line in this area here when it wasgoing in and… resistance to it, particularly the elevated component (V3).4.2 Rental Housing and ImmigrationThe concentration of low incomes in the corridor can be explained to a certain degree by thepresence of a variety of forms of affordable rental housing occupied by low income people.Rental housing in the corridor is found in purpose-built low-rise apartment buildings and towers,secondary suites in houses, or condominium units rented out privately by an owner. In themajority of interviews it was 3-4 storey, wood-frame, purpose-built rental housing that was citedas explaining the presence of particularly low incomes in the corridor. The Expo SkyTrain Linecarries far more passengers than any other line in the region’s transportation network (TransLink2007) and regardless of whether the SkyTrain Line is a cause or just a correlate of lower propertyvalues, housing costs, and incomes, rapid transit plays an important role in the lives of low-income people.  They have a higher propensity to rely upon public transit, so having access totransit is an important consideration in residential choice.Public transit for low income people and for young people is definitely significant; andnew immigrants. Across the region that’s the case (S1).We know renters have a much higher propensity [to use public transit] than owners… Forsome people it also depends on household income and auto ownership. Those are the twokey variables. If you have lower household income, or if you don’t own a car, well it50doesn’t matter how old you are. You’re going to need at least to have access to transit(R2).I think that what happens is having Skytrain attracts people who maybe can’t afford tohave a car and they're going to be renting cheaper housing… It also attracts somestudents… because it’s right on Skytrain (V2).The presence of affordable rental housing occupied by vulnerable recent immigrants emerged asa major theme; particularly in areas that are well served by public transit. Refugees are closelyassociated with the prevalence of low incomes in certain places.[The area] is providing affordable housing to people who have recently arrived, and asyou say, it’s close to transit. So when you're getting your feet on the ground it’s prettyimportant (B2).One of the things I've noticed, and people have talked about with Collingwood Village…I think it does attract a number of immigrants because it’s right on Skytrain (V2).CEJ: How important is public transportation to the [recent immigrants andrefugees] you serve?B5: It's huge. They're all on public transportation.CEJ: When someone asks you for help to find new housing do they ask foraccess to good public transit?B5: Totally.[Finding affordable housing for newcomers] is probably one of the biggest challenges…Some of the other factors that come into consideration is being near their ethno-culturalcommunity, or being near other newcomers is really important, so having that supportnearby would also become a factor. And then access to transportation. Most people won’thave cars (R1).51It was also noted by informants that the number of vulnerable recent immigrants and refugeesarriving in the City of Vancouver has been declining. Ley and Lynch (2012) show how importantthis corridor was for recent immigration in 2005, and figure 8 below shows that the corridorcontinues to be an important landing site for new immigrants, but recent immigration does notbecome concentrated until close to the eastern boundary of Vancouver, then continuing intoNorth Surrey.Figure 8 Recent Immigrants in Vancouver CMA. Source: Hiebert & Jones (2014) (reproduced withpermission)An immigrant services professional with an agency that finds housing for new immigrantscommented that in Vancouver, apartment buildings and single detached houses tend to be52interspersed, whereas in suburban developments there is often a separation between districts ofapartments and of houses. The concentration of low incomes becomes visible at the CT levelwhen there is a massive grouping of affordable apartments.I think in some places the visibility and the concentration becomes visible because thereis this massive grouping of affordable apartments… it’s block after block… fourbuildings with another block of three or four buildings, and then another one… So I cansee the newcomers there in part because I can see the apartment buildings (R1).One group of newcomers that arrives in Canada with very limited resources are governmentassisted refugees (GARs). GARs spend the first two weeks after landing in Canada in thedowntown Welcome House operated by the Immigrant Services Society of BC (ISSofBC).Amongst other services, ISSofBC finds initial housing placements for GARs, and they record thepostal code of where they successfully find housing. Figure 9 shows GAR initial housingplacements for the years 2005-2009.53Figure 9 Metro Vancouver GAR Arrivals, 2005-2009. Source: Brunner & Friesen (2011) (reproduced withpermission)The low income SkyTrain corridor is clearly evident, as is the diminishing role that Vancouverplays in receiving refugees. Finding appropriate housing for GARs can be challenging as theyhave limited incomes upon arrival and tend to have larger families. Although the number ofGARs arriving every year is unlikely to have a significant effect on average incomes at the CTlevel, the initial housing placements of GARs acts as an indicator of the location of the region’saffordable rental family housing. Figure 10 shows a series of maps I prepared for ISSofBC toextend the analysis through 2010-2013. The corridor continues to be visible as is the lowernumber of GARs finding housing in Vancouver.54Figure 10 Metro Vancouver GAR Arrivals, 2010-2013. Source: Friesen & Daily (2014) (reproduced withpermission)Figure 11 maps the residential location of refugees in 2011; the corridor is evident, and a handfulof CTs within the corridor can be seen as being of particular significance to the regionaldistribution of refugees. In these areas refugees strikingly represent at least 5-10% of the totalpopulation.55Figure 11 Refugees in Vancouver CMA. Source Hiebert & Jones (2014) (reproduced with permission)4.3 A Corridor of Contingent DevelopmentThe SkyTrain route is a corridor of contingent development in addition to being a low-incomecorridor where affordable rental housing is often occupied by vulnerable recent immigrants forwhom access to public transit is particularly important. In each of the four municipalities that thecorridor runs through, high density developments can be found near SkyTrain stations, withseveral high-density projects currently under construction or being proposed. A commonargument in support of these projects is that increased density near public transit infrastructure isa sustainable way to accommodate population growth. Within the next 20 years it is expected56that several neighbourhoods around SkyTrain Expo Line stations will see significant growth intheir populations and in the number of high-density developments. However, the process ofdevelopment is contingent upon a number of factors and the presence of rapid transit does notautomatically mean that development will take place. It is the interaction between rapid transitinfrastructure, municipal policy, and the market which determines whether or not developmenttakes place at a given SkyTrain station (Coriolis 2013).The bottom line… is that rapid transit is an important, but not dominant factor forinfluencing high density development. You also need support of municipal policy, and asupportive market as well (R2).4.3.1 VancouverClose to the SkyTrain corridor, the Vancouver neighbourhood of Mount Pleasant used to be alanding site for recent immigrants, but it is currently undergoing gentrification, making itdifficult for new arrivals to find housing there. However, some affordable private rental housingremains in the form of secondary suites in houses and there is a considerable stock of purpose-built, low-rise rental apartments.Mount Pleasant used to be very much a place where new immigrants could come and getlow income housing, seniors were able to find low income housing but over the last fiveto ten years this community has really changed a lot and it’s certainly being gentrifiedand working its way [eastward]… So we’ve got the hip, the artistic, the youngprofessionals in the neighbourhood that own these beautiful houses around us, but insidethose beautiful houses there are basement suites and on the north side of the street thereare a lot of apartments that are still reasonable rents and so that actually still means thatwe do have some families and some other people who are low income (V1).57Ground is currently being broken for a controversial tower development in Mount Pleasant. Thisproject initiated by the Rize development group features a 19-storey tower that will dwarf otherbuildings at the intersection of Main Street and Broadway. The proposed height drew“considerable community opposition” which was debated at council during an exceptional sixevenings of public hearings (Lee 2012). Council approved this project despite the opposition as itwas seen as necessary to plans for denser transit-oriented housing along the Broadway corridor.Vancouver Mayor Gregor Robertson’s position was that, “We need to think in the long term andensure we have lots of housing along the transit routes” (ibid).To the east of Mount Pleasant, and on the SkyTrain alignment, the neighbourhood of GrandviewWoodland has one of the largest proportions of renters in Vancouver. Rental housing in the areacomes in many forms: low-rise apartments, secondary suites, and compartmentalized heritagehouses.It actually gets built out in a number of different ways, you know, two-storey, three-storey typically… apartments and that’s a part of that affordable stock there. You also getwithin the neighborhood… [a] heritage core area (V3).An ‘emerging land use directions’ map was released by the City of Vancouver which proposedzoning for a 36-storey tower next to the SkyTrain station at the intersection of Broadway andCommercial  in Grandview Woodland (Vancouver 2013). The proposal came as a shock to manyand drew criticism from local residents for proposing a level of density and height that wasinconsistent with their vision of the neighbourhood (Cole 2013). In the face of communityopposition the City backed away from the plan and established a citizens’ assembly to re-thinkland use decisions for the area (Metcalfe 2013).58In these two Vancouver neighbourhoods, decisions which from a planning perspectivereasonably aim to introduce high-density developments near transit have come into conflict withthe priorities of local residents.To me, Broadway and Commercial is a significant SkyTrain station surrounded by singleand 2 family dwellings [interviewee laughs], so from a planning perspective in terms ofdensification [this doesn’t make sense] (B1).We expect… another 140,000 people in a few decades… and so when we think abouthow to accommodate that growth, there’s a number of different choices you could make.You could map out an even distribution. You could carry on [with] the developmentpatterns that’s already taking place or you can… adopt a more sophisticated approachwhere you look to the areas where there’s significant public investment, where there’ssignificant transit. Now do we want more people driving in cars around town? No, ideallywe’d like to give them options that allow them to make use of sustainable transportation.Do we want to try and maximize the public investment in the social and culturalinfrastructure that we’ve got? Absolutely, and so… this helps to set the table for someland use [decisions]. When we look ahead... over the next few decades in Vancouver, Ithink we see one of the significant opportunities is the presence of our SkyTrain and rapidtransit infrastructure… Broadway and Commercial seems to have on that basis alone,some pretty significant opportunities around it… This is already a transit orientedcommunity in that sense…That’s not going to change and we’ve got hard choices tomake when we look down the road into the future about how we want to accommodatepeople (V3).The neighbourhood of Joyce Collingwood in South-East Vancouver is just west of the boundarythat separates Vancouver from Burnaby. It is home to one of Vancouver’s first transit orienteddevelopments; a high-density area known as Collingwood Village, next to the Joyce SkyTrainstation. Concert Properties owns and operates several purpose built rental building towers, and59there are several condominium towers in which many units are rented privately by owners.Secondary suites are also an important part of the area’s rental housing stock. The purpose-builtrental tower apartments in Collingwood Village and privately rented condominium units in stratatitle towers are home to large numbers of recent immigrants. Although data on the turnover-ratein the rental towers is unavailable, one informant believed that residential turnover among recentimmigrants is high.There are a lot of rental suites that are in houses. We're expensive and so people have toput in a rental property…  There are condominiums, a lot of them rented out. So you dofind that there is quite a movement of people… an immigrant might come there and thenmove within a year or two. Then somebody else is coming in…(V2).In this already dense neighbourhood, high-density development was seen as necessary for thecontinued growth of the city.We have to densify in Vancouver. There’s just no question in my mind… What'shappening in Grandview Woodlands and what was happening around Main Street, aswell, with the Rize development, people don’t want to see that density. They wantaffordability, but they don’t want to see that density (V2).4.3.2 BurnabyMetrotown is a regional town centre with a SkyTrain station, anchored by a major retail mall andone of the region’s largest concentrations of suburban office space. The Edmonds town centre ofBurnaby, also with a SkyTrain station, is a well-established landing site for recent immigrantsbecause the stock of low-rise purpose-built rental apartments provides newcomers withopportunities to find affordable housing in larger units. Informants in Burnaby expressed a clearunderstanding of the link between affordable housing, immigration, public transportation, and60the concentration of low-income people in particular places. Burnaby is second only toVancouver in terms of the absolute number of aging purpose-built rental units and along theSkyTrain Line in Burnaby there is a large stock of purpose-built, low-rise rental apartmentbuildings which predate the construction of the rapid transit line (Coriolis 2012b). Many of theseunits are affordable and large enough for families with limited incomes. In both Metrotown andEdmonds there is concern that this housing stock is aging.The purpose-built rental was there way before the Expo Line came in. The fact that it’sall rental… it’s going to be for lower income people… Because you need a landing areawhen you're recently arrived, you're going to go to the purpose-built rental (B2).In the case along the corridor in Burnaby, I think it was a lot of three-storey walk-upapartment buildings that have gradually over the years deteriorated so that they’vebecome really the lowest form of housing… we've retained two- and three- andsometimes four-bedroom apartments along the corridor that are almost non-existentanywhere else in the Lower Mainland, so families move into those, particularly refugeefamilies will move into those because they're going to have more space in thoseapartments for families that are larger (B4).Metrotown is expected to receive approximately 30,000 new residents during the next 20 yearsand several condominium towers are currently under construction to accommodate some of thisgrowth (Smith 2013). Burnaby Councillor Colleen Jordan says, “Part of our responsibility to theregion is to allow more places for people to live… So those are the places where we havedesignated for high density: density around transit, housing around transit” (ibid). There areseveral high-density developments in the Edmonds town centre of Burnaby with more plannedfor the near future. Highgate Village is a recently completed development of four residentialtowers anchored by a village-style mall, and Cressey Developments has plans for a three tower61development at the intersection of Kingsway and Edmonds Street. These towers will stand at 28-,31-, and 37-storeys and planning is in process for the redevelopment of a huge 48-acre site to theeast of the Edmonds SkyTrain station (Burnaby Newsleader 2014).At the centre of Burnaby’s development boom are policies that encourage transit-orienteddevelopment. Developers are investing heavily in Metrotown due to high residential demand andbecause the City of Burnaby has good working relationships with developers. According toMichael Ferreira, real estate expert and co-owner of Urban Analytics Inc., “Burnaby is anattractive market because of its central location… with two Skytrain routes going through it”(Jang 2013). Bob Ransford, an urban development consultant praised Burnaby for doing, “a greatjob developing density around the SkyTrain stations, more so than what Vancouver has done”(ibid).  According to Hani Lammam, vice president of development and acquisitions for CresseyDevelopment Group, “Burnaby is a good place to do business because you know what you’regetting into… You know what it’s going to cost you” (ibid).Public opposition to development in Burnaby is minimal and a project that will make Burnabyhome to the highest building in the Vancouver region was approved with “barely a whimper”(Smith 2013). However, some key informants were concerned that densification in Burnaby willlead to the displacement of low-income families.People are having to move out because of housing. Housing is a definite challenge andissue. We have developers coming in, tearing down the housing stock and buildingtowers that aren’t going to be affordable for families. One example of that is a tower…that was rental housing and from what I gather… (the developer) provided those familieswith a buy out of money for rent for a couple of months so they could get relocated, but Iwould guess that those families didn’t relocate in Burnaby, that they had to relocate62maybe in Surrey or somewhere else. So those towers are going up but they're notaffordable towers (B3).4.3.3 New WestminsterFor a geographically small municipality, New Westminster has a remarkably large stock of agingPBRH; second only to Vancouver as a proportion of overall housing (Coriolis 2012a). In theuptown area of Brow of the Hill, which is close to the Edmonds area of Burnaby, there is aconcentration of wood-frame, three-storey walk-up PBRH. These older apartments are affordableand tend to be larger than basement suites and the new rental housing stock, which makes themattractive to recent immigrants with children.We have close to 40% of the population that are in Brow of the Hill in uptown that areimmigrants. When you look at those areas, we also know they have higher proportions ofnew immigrants, higher proportions of government-assisted refugees and the reason is ina lot of cases, some of that rental housing is larger... It’s all three story walk-ups, it’s allpurpose-built rental… In a lot of cases, that rental housing is a lot older so they can’tcharge as much for rent (N1).In New Westminster there has been a return of capital investment and development to thedowntown area around the SkyTrain stations. As BC’s former capital city, New Westminster hasa long-established urban downtown that is atypical of other Vancouver suburbs. NewWestminster Councillor Jonathon X. Cote credits, “The urban environment,” for the return ofcapital to the downtown because it, “is actually becoming a value: being able to walkeverywhere, being able to take public transit, being able to jump on a SkyTrain to go downtown”(Smith 2012). As of August 2013 there was such a high level of interest from the private sectorthat not all proposed projects could be approved.63We’ve got seven potential proposals [for downtown next to a SkyTrain station]… thedifficulty there is there’s just too many competing towers. Where is all that interestcoming from? Part of that is just the convenience of being close to the SkyTrain and[therefore being able to] command higher sales prices (N1).But it’s not just the urban environment and the convenience of SkyTrain which is attractingdevelopment; there are policies in New Westminster that are designed to attract development tothe downtown, particularly around the SkyTrain stations.The city offers a whole bunch of incentives to develop there. If you develop near theSkyTrain, you reduce the parking allotments and you’re looking at underground parking.That can be up to $40,000 per stall and we’re actually saying you don’t need as much ifyou’re near a SkyTrain (N1).4.3.4 SurreyParticularly low income people are attracted to North Surrey because they can find a largerapartment unit that is more affordable than elsewhere in the region. Recently built condominiumtowers also present an opportunity for home ownership to first-time buyers with lower incomes.There are huge complexes of PBRH in North Surrey. Built mostly in the 1970s, this aging stockof affordable housing plays an important role in providing housing options to those near thebottom of the region’s housing market.I think it’s more about rental housing, I would guess it’s more about rental housing thananything... I think why particularly low income people are attracted to [North Surrey] isthe cost of housing… It’s that four-storey, wood frame [PBRH], the complexes here arehuge... At this point some of them are very run down… In terms of ISS placing GARs inhousing in the region, this is where they can find larger apartment units that are morereasonably priced than elsewhere in the region (S1).64There are concerns that the concentration of refugees in apartment complexes in North Surreyhas created an ethnic enclave that is excluded from the rest of the city, making integration morechallenging.Where the concern is coming is on the refugee enclaves. It’s creating challenges up in theGuildford area in particular. Part of it is they are settled, often many families in onecomplex, so you’ve got a number of families, which I would imagine there’s pros andcons. The pros being the support that they would offer each other; the cons being they'rehaving more challenges integrating. I know I've heard from a youth worker that works inthat area [mainly with refugees]… her comment to me was the young people she workswith… don’t have any Canadian friends. They're afraid to even approach anyone atschool because they feel like such outsiders (S1).A transformation is taking place in the City Centre area of Whalley in North Surrey. In order tocreate a downtown core the City has invested heavily in this area. Simon Fraser University hasopened a new campus, a new library designed by Vancouver architect Bing Thom has been built,and a new Surrey City Hall is under construction. At least 10 more towers are expected in thenext decade along with other projects that aim to create a dense, urban core around the SkyTrainstations in North Surrey (Sinoski 2013). The population of the area is expected to double, anddensity is expected to reach a level comparable to two of the region’s most developed areas,Yaletown and Metrotown. This is generally welcomed by the community.I was expecting that there would be people in that neighbourhood who were upset withthe change. I think that if this were happening in many areas of Vancouver there wouldbe a lot of people coming out upset. There wasn’t that at all… They're seeing thisimprovement; they're seeing the new library… For the most part there is a lot of support(S1).65As there is a substantial land base in North Surrey that has yet to be developed at even mediumdensity, development pressure around SkyTrain stations is focused on areas of single familyhouses.We’re not getting any pressure to demolish any of the old rental stock… there’s so muchland… so if a developer is assembling land for redevelopment , they're going after singlefamily lots at this point (S1).4.4 Issues Associated with Development: Land Speculation and DisplacementAlthough it is apparent that densification in the corridor is varied in its implementation, form andthe community’s response to it, a common issue emerged; land speculation as a consequence ofthe expectation that greater densities will be allowed in transit corridors. Some landowners aresimply holding onto properties with the expectation that future land values will surpass currentland use designations. This expectation acts as a disincentive to invest in the maintenance ofbuildings, which is a particular problem for rental properties when landlords weigh the cost ofproper maintenance against the potential to sell for redevelopment.There have been a couple of apartment buildings… on the north side of [Broadway inMount Pleasant] that have been recently renovated. However, if you have a walk in thatarea you can really see that there are some that are really challenged. I think that theyhave lost their prime, they aren’t always as well maintained as they should be. There aresome beautiful places, but you notice which ones which would be ready for majorrenovations and I think the challenge is that because of the development of the area Ithink that some owners would rather develop as opposed to renovate (V1).There is some affordable rents [in Burnaby] and they're not subsidized rents, but they'reaffordable so that families can still live in them. But the apartments are getting old. Itmakes sense that whoever owns that land is going to sell it to a developer and make66money and that’s good for who owns the land, but [not] the people who are renting…(B3).We've got a few situations [in Edmonds] where clients are living in these buildings,landlord is trying to sell it; building is falling apart, literally. Like there might be a deckon it that our clients cannot go out onto because it's literally unsafe to go out onto.Landlord’s planning on selling the place [so they won’t do maintenance] (B5).There’s all these towers that’s going in. At some point, [a landlord is going to make a]decision, “Do I want to invest in my building and try to prolong its life into 30 years oram I just going to leave it because look what’s happening down the block?” (N1).What [Surrey City Centre] is experiencing is the classic of what happens as an area isgoing through [a change in density]. There’s a lot of vacant lots, and some of the issuesthat go along with people holding on to pieces of property and not maintaining theproperty. I mean a lot of investors who are property owners, but aren’t invested in thecommunity… So there’s been issues with grow ops and crack houses and things like that.I think that’s the challenging side of an area in transition [from low to high density] (S1).Land speculation is a consequence of policies designed to encourage high-density developmentnear rapid transit, as is the threat of redevelopment to affordable rental housing near SkyTrainstations. For vulnerable groups living in affordable rental housing near SkyTrain stations, thedemolition of aging PBRH could lead to displacement further from the City of Vancouver.I think transit-oriented developments are great, but they often come with gentrification.And we’re seeing that challenge right now… if you look at how many of those two- tothree-storey walk-ups there are now compared to 5 years ago, my sense it that thenumbers are declining, because people can rip them down, put up a high-rise. [Then]rents go up because it’s more of a desirable location (R1).67CEJ: Does development pose a threat to the stock of affordable housing aroundSkytrain stations?B1: I can be up front about that and say that it does. The plan reviews aren’tcomplete yet, but there is a very good chance that those areas will be re-designated for high density development. Certainly higher than 3 and 4storey wood frame.More towers are going up. And that stock of housing that used to be affordable for ourrefugee clients has really been decreasing… People are just moving south. People aregoing to Surrey. That's basically where the refugees have to settle… they're going wherethe housing is and that's where it is and that's where they can afford to live (B5).Along the SkyTrain Line there is a coexistence of affordable rental housing occupied by lowincome people, and high density development near SkyTrain stations. Conflicts arise between thetwo when policies that encourage high-density developments lead to the deterioration ordemolition of affordable rental housing, which can cause the displacement of vulnerable peoplenear the bottom of the housing market who rely upon access to public transit for mobility.CEJ: Do you think there’s an unavoidable tension between access to publictransportation and affordability?V3: Well wouldn’t that be like the irony of it all, of course, in trying to do thismore affordable, environmentally sustainable [development]… I think thekind of the missing component to that discussion and the one that’s likethe big elephant in the room, is that transit should be something that iseasily accessible for all folks… Right now we have so many componentsof our transit system that are at capacity or near to capacity and so we cantalk all we want about the importance of doing these sorts of things, andyou’re right, you hit this sort of strange supply-demand thing where if we68actually don’t have a whole pile of transit everywhere then you have it in amore specialized series of routes or locations… you do run the risk of thatsort of thing, so my answer’s a bit hypothetical but, no, I don’t think thereshould be a tension or link there because ideally we’ve got enoughtransit...However, funding for transit has been a complicated issue of late (CTV News 2014) and it isbeyond the mandate of the regional transit authority, TransLink, to consider the effect that rapidtransit has on affordable housing. Although TransLink is aware that frequent transit networks,“May increase the demand for and therefore market value of land in proximity to the network,providing incentives for the highest and best use of the land for tax revenue” (Walker et al.2009), it is the responsibility of municipalities to ensure that affordable housing near transit isprotected.When it comes to PBRH, we would see that as a municipal and local governmentresponsibility. From our perspective we don’t get into that level of detail, so if we havesome alignment, we are not going to assess the impact of PBRH because… all amunicipality needs to do is say we’re going to put a policy in to not touch this or have nonet loss of PBRH so that’s outside of our control, and at the end of the day there’s alldifferent forms and types of development that are transit supportive, so it’s up to themunicipality to do that. The main thing is that we’re getting the mix of uses and thedensity that support the transit ridership or the future transit ridership (R2).Therefore, the management of the tension between affordable housing and access to publictransit falls into something of a regulatory vacuum, with no authority responsible for makingsure that transit is available to those struggling in the region’s housing market with the greatestneed for access to transit.694.5 ConclusionThe Kingsway SkyTrain corridor was established in the colonial era and has shaped regionalmigration patterns. Lower property values are one consequence of this being a long-establishedrail, tram, and automobile corridor, and the development of rental housing provides affordableoptions to those with lower incomes. Whether or not the SkyTrain Line itself can explain theemergence of low incomes in the corridor is therefore unlikely, as the route of the SkyTrain maybe more a consequence of an existing rail right-of-way and a lack of effective political resistance.Regardless, there is a considerable stock of rental housing in a variety of forms in the corridor,and some of this rental housing provides opportunities to low-income people near the bottom ofthe region’s housing market.Areas of affordable rental housing that are well served by transit are particularly important forvulnerable recent immigrants living in the region, and this is especially true for refugees. Datafrom ISSofBC and the 2011 Canadian National Household Survey show that the low incomecorridor is an important site for initial housing placements of GARs, and for the long-termresidential location of refugees, with suburban municipalities receiving an increasing share ofrefugees over time (Brunner & Friesen 2011; Friesen & Daily 2014; Hiebert & Jones 2014).Although it is unlikely that the number of refugees in a CT could have a major effect on averageincomes, where refugees are able to find housing is a proxy for locating the Vancouver region’smost affordable rental housing stock. And as public transportation is particularly important forrefugees, it also serves as an indicator of where affordable rental housing is well-served bytransit.70There are examples of development, often at high densities in every municipality in the corridor.As the population of Metro Vancouver is expected to grow significantly in the next 20 years,accommodating population growth through high-density development has become a widelyaccepted policy option, as long as high-densities are well served by public transit. This is not tosay that there will be high-density developments at every SkyTrain station in the corridor, forthis is a process that is mediated by a variety of factors. The presence of rapid transit isimportant, but so are municipal policy, private investment, and community response to high-density rezoning.In the City of Vancouver, two traditionally low-income neighbourhoods with a large stock ofaffordable rental housing have seen proposals to allow for high-density development in recentyears. From a planning perspective, adding density at two key intersections that are well servedby public transit is an effective way to accommodate growth, but public reaction to theseproposals was largely negative. Although the Rize development in Mount Pleasant was approvedby Vancouver City Council despite opposition, an emerging land use directions document forGrandview Woodland was tabled for further community consultation after public outcry overproposed increases to density. Both Mount Pleasant and Grandview-Woodland have a history ofimmigration, but recent immigration in these neighbourhoods is no longer as significant as itonce was. Recent immigration in Joyce-Collingwood continues to be significant, particularly inthe high-density developments of Collingwood Village, where densification was seen asinevitable.Low income areas in Burnaby were generally explained by the presence of rental housing, andparticularly low incomes were strongly associated with a stock of low-rise PBRH that pre-dates71the construction of the SkyTrain Line. This aging housing stock is important to recentimmigrants and refugees with larger families, as the apartments are affordable and large; apairing that is becoming increasingly difficult to find in the Lower Mainland. In anticipation ofpopulation growth, the City of Burnaby is increasing density near SkyTrain stations. Thesepolicies are welcomed by the development community and public opposition is minimal, butthere are concerns that increased density is putting redevelopment pressure upon affordablePBRH which could lead to the displacement of vulnerable families.In New Westminster there is a stock of aging PBRH in the Brow of the Hill area that provideshousing opportunities to vulnerable recent immigrants. There is a huge amount of developmentinterest from the private sector and the City of New Westminster is promoting high-densitydevelopment in the downtown core near SkyTrain stations.North Surrey possesses a huge stock of concentrated, aging PBRH in which refugees are able tofind larger units at a lower price. There are concerns that this may be creating an ethnic enclavewhich inhibits the integration of young people into the community, but there is little concern thatdisplacement could happen soon. In an effort to create a downtown core the City of Surrey istransforming the City Centre of Whalley into a high-density transit oriented community. As thereis a large land base of low density housing that can be assembled by developers, there is littleredevelopment pressure on the aging stock of PBRH at present.Rental housing, vulnerable recent immigrants and high-density development must be consideredtogether in relationship to the SkyTrain Line in order to understand the processes ofneighbourhood change taking place within the corridor. The region-wide policy discourses whichfavour high-density developments near public transit routes make sense from a planning72perspective, but there are issues associated with the expectation that land values near transit willincrease because of these policies. For one, land speculation is causing problems in many placesas landowners are not investing in the proper maintenance of their properties. This is a particularproblem when landlords of aging PBRH weigh the cost of maintenance against the potential tosell for redevelopment. Another consequence of high-density development in the corridor is thedisplacement of vulnerable residents, forcing them to look for affordable rental housing whichmay not be as well served by transit. The end result verges on the ironic: for a variety of reasonsaffordable rental housing was developed along a rail right-of-way which was established longbefore the SkyTrain Line was built in the early 1980s; access to rapid public transit made theseareas more attractive and as the population of the region is expected to continue to grow, high-density developments near SkyTrain stations make sense from a planning perspective; policieswhich aim to encourage higher densities put redevelopment pressure upon the stock of agingPBRH which could lead to the gentrification of these areas and the displacement of those whohave the greatest need for access to transit. This type of situation is superbly critiqued by Grube-Carvers & Patterson (2014):If transit is not made accessible to those populations who would receive the greatestmarginal benefit from its use, then it is not fulfilling its role of increasing the equity andaccessibility of urban spaces.73Chapter 5: Neighbourhood Change in the Low-Income CorridorIn this chapter I further narrow the focus of my research to the third scale of analysis; twoneighbourhoods that are within the low-income corridor. The primary research method at thisscale was a series of four focus groups with residents of the two selected neighbourhoods. Thegoal of this method was to gain insight into income polarization and neighbourhood change fromthe perspective of those who live within the corridor. A variety of indicators were used in theprocess of selecting the neighbourhoods. I developed measures of recent immigration and theresidential distribution of refugees, which were then layered with indicators of low income andrental housing disadvantage from census sources. Those areas in which several of theseindicators proved to be significant were selected for focus groups.In this way the neighbourhoods of Maywood and Richmond Park in Burnaby were selected. Itwas found that these neighbourhoods possessed many of the qualities that were found to besignificant in chapter 3. Both neighbourhoods have seen a significant increase in the populationof visible minorities since 1986, they are important sites of recent immigration, the majority ofthe housing stock are apartments occupied by renters who were spending a large share of theirincome on housing, and the incidence of low-income families in these neighbourhoods is muchhigher than the Burnaby average.Focus group participants were recruited by staff at the South Burnaby Neighbourhood House andthe Edmonds office of MOSAIC BC. The characteristics of these participants fit well with theindicators that led to the selection of these neighbourhoods. Findings from the focus groups were74supported by a review of policy documents and rezoning applications and an interview with anelected government official.The key findings of chapters 3 and 4 informed the selection of these neighbourhoods, therecruitment of participants and the line of questioning pursued in focus groups. Focus groupsproved to be an excellent method for gaining a textured account of these neighbourhoods whichboth confirmed and confronted expectations.5.1 Neighbourhood Selection ProcessIn order to gain insight at the neighbourhood scale it was determined that the best researchmethod would be to conduct a series of four focus groups with residents of two neighbourhoodswithin the low-income corridor. The neighbourhood selection process was guided by a series ofindicators from a variety of external sources and from my own research at the regional scale. Theneighbourhoods that I selected are both in Burnaby and they are identified by the City ofBurnaby as Maywood and Richmond Park.One of the original indicators in this research project is average household income as mapped bythe Cities Centre (Ley & Lynch 2012). In figure 12 below, CTs in Maywood and Richmond Parkare amongst a handful of CTs in the region which are defined as very low income status in 2006.75Figure 12 Very Low Average Household Income in Maywood and Richmond Park. Source: Ley & Lynch(2012) (reproduced with permission)In the OLS and logistic regressions in chapter 3, recent immigration was a statistically significantand meaningful variable in determining the average income of a CT and was likely to be high inareas that experienced income decline. I refer to figure 13 below which maps recent immigrationas derived from 2011 National Household Survey data. As in Ley & Lynch (2012) the corridor isevident and one of the CTs in Maywood has one of the highest concentrations of recentimmigrants in the region.76Figure 13 Recent Immigrants in Maywood and Richmond Park. Source: Hiebert & Jones (2014) (reproducedwith permission)Figure 14 shows the residential location of refugees (landing 1980-2011) in 2011. In this case thecorridor is not as well-defined, but some of the CTs which comprise Maywood and RichmondPark are amongst the most significant in the region.77Figure 14 Refugees in Maywood and Richmond Park. Source: Hiebert & Jones (2014) (reproduced withpermission)The data for refugees landing in Canada from 2006 to 2011 are particularly informative, and it isevident in figure 15 that CTs in and around Maywood and particularly in Richmond Park are keylanding sites for recent refugees to Canada.78Figure 15 Recent Refugees in Maywood and Richmond Park. Source: Hiebert & Jones (2014) (reproducedwith permission)Figure 16 maps a rental housing disadvantage index based on the 2006 census, which identifiedCTs in Maywood and Richmond Park as having the highest rental housing disadvantage in theregion. This is a distinction shared only with one CT in Richmond as well as Vancouver’s wellknown district of poverty, the Downtown East Side (DTES).79Figure 16 Rental Housing Disadvantage in Maywood and Richmond Park. Source: NCRP, Univ. Toronto(2014) (reproduced with permission)Although the boundaries of neighbourhoods as defined by municipalities do not align perfectlywith CT boundaries, I am confident that there is sufficient overlap of the indicators above tohave selected Maywood and Richmond Park as sites in which to conduct focus groups.5.2 Community Profiles of Maywood and Richmond ParkTable 4 Community Profiles of Maywood and Richmond ParkPop. Growth(1986-2006)Immigrants English spokenat home (1986)English spoken athome (2006)Single and twofamily homesTownhouseand low-riseapartmentsHigh-riseapartmentsRenters Renters spending morethan 30% of monthlyincome on housingPrevalence of LowIncome FamiliesBurnaby 43% 51% 90% 62% 45% 36% 19% 39% 43% 22.10%Maywood 63% 66% 94% 41% 1% 49% 50% 65% 51% 38.60%Richmond Park 40% 59% 95% 55% 29% 48% 23% 70% 51% 35.90%Source: City of Burnaby Neighbourhood Profiles, 2006 Census80Table 4, above, shows that the City of Burnaby, Maywood and Richmond Park have allexperienced some remarkable changes in the 20 years between 1986 and 2006. All areasexperienced significant population growth, which has largely been the result of immigration asthe majority of the populations of Burnaby, Maywood and Richmond Park are now immigrants.The drastic reduction in the percentage of people that speak English at home reflects thechanging trends in country of birth for immigrants to Canada. We can see that the characteristicsof the housing stock in Maywood and Richmond Park differs significantly from that of Burnabyas a whole. Whereas the dominant form of housing unit in Burnaby is single- and two-familyhouses, apartments are the dominant form of housing unit in Maywood and Richmond Park. Itshould be noted that the number of apartments in a high rise building is far greater than that oftownhouses or low rise buildings, so the number of high rise properties is significantly less thanthe number of townhouses and low rise apartment properties. It is far more common for residentsof Maywood and Richmond Park to be renters than for Burnaby as a whole, and renters in theseneighbourhoods are more likely than not to be spending more than 30% of their monthly incomeon housing, therefore experiencing housing stress. In addition to this, families in Maywood andRichmond Park have a significantly higher incidence of low income than for Burnaby as awhole.As almost half of all housing units found in both Maywood and Richmond Park are apartmentsin townhomes or low rise buildings, it begs the question as to why there are so many units of thistype in these neighbourhoods? The presence of this housing stock can be explained by fourfactors:811. Planning decisions made in the 1950s favoured the development of apartment buildingsin town centres to meet growing residential demand (Beasley 1976).2. During the time that much of Burnaby’s purpose-built rental housing stock wasdeveloped, there existed a number of tax incentives at the federal and provincial levels ofgovernment to support the development of market rental housing (Carter 1997).3. Low-rise PBRH in Burnaby has been protected from strata conversions since 1974 whenthe City of Burnaby adopted a moratorium on all such conversions (Burnaby 1974).Although this moratorium is still in effect, it will be shown below that in no longerapplies to in Burnaby’s four town centres.4. The zoning for many of these low-rise rental buildings has not changed since they wereoriginally developed (B1); therefore it has not been possible to redevelop these sites athigher densities.5.3 Focus Groups in Maywood and Richmond Park5.3.1 RecruitmentI relied upon third-party recruitment to gather participants for each focus group. Two focusgroups were conducted in Maywood, and two focus groups were conducted in Richmond Park.In Maywood, participants were recruited by South Burnaby Neighbourhood House, and inRichmond Park participants were recruited by the Edmonds office of MOSAIC BC. Individualshad to have prior contact with one of these organizations in order to be recruited. For the twofocus groups in Richmond Park an interpreter was present in order to ensure understandingbetween myself and participants. English was a second language for the majority of participantsand I have decided not to edit their words unless absolutely necessary for clarity.825.3.2 Selected Demographic Characteristics of Focus Group ParticipantsTable 5 Selected Demographic Characteristics of Focus Group ParticipantsAll participants were immigrants, and most self-identified as a visible minority. Mostparticipants had lived in their neighbourhood for 3 to 9 years, and few had lived in theirneighbourhood for 10 years or more. Only one participant was working full time and allparticipants but one were renters. I cannot claim that participants are representative of theirneighbourhoods, but they do bring a perspective that is particularly salient in the context of thisresearch; that of mostly visible minority immigrants, from a variety of ethnic backgrounds, whoare renting apartments on limited incomes.5.3.3 Major Insight from Focus GroupsWith the measures of low income, rental housing disadvantage, refugee concentrations and highrates of recent immigration, I must admit that going into the focus groups I expected participantsto express a perception of their neighbourhood that was generally negative. However, I wasMaywood Richmond ParkNumber of participants 13 13Male 1 8% 5 38%Female 12 92% 8 62%Average Age 39 46Age Range 16-58 25-70Immigrant 13 100% 13 100%Visible Minority 9 69% 13 100%Resident of neighbourhood for less than 3 years 3 23% 3 23%Resident of neighbourhood for 3 years or more 9 69% 8 62%Resident of neighbourhood for 10 years or more 1 8% 2 15%Working full time 1 8% 0 0%Working part time 5 39% 6 46%Participant or member of family owns home 1 8% 0 0%Average number of persons per household 3.9 4.683pleasantly surprised that participants expressed unanimous appreciation of the neighbourhoods.Some of the major themes of appreciation for the neighbourhoods were their walkability, anattractive and bustling street environment, easy access to a variety of services, and safety.I moved here to south Burnaby, because to raise the kids there is lots of things which areconvenient for the kids. There is the shopping mall and library and swimming pool and soconvenient for the kids… And for me because I don’t have a car it is easy for me to getaround.There are a lot of resources… not too far from your house. There are settlementworkers… there is a newcomers’ centre for children and families which is really good foreverybody if you need information.[T]he schools are very close. You can drop your kids walking; they can walk to school.It’s beautiful, it’s quiet, sophisticated. With the passage of time it becomes better. I like itvery much.The area is amazing. It’s beautiful, clean, safe. Everything here is more than what weimagined.We go out at night, we see neighbours… we want our community to be open, to bemixed.Like I don’t feel scared or threatened by anything… I have been here for 9 years now. Ifeel very safe walking at night like when I am taking the bus; I’m not scared at all.It’s a good area. It’s a very safe area. Sometimes we don’t even have to lock [our doors].Another theme that was important to participants was the welcoming nature of the community.This welcoming character was due in part to the presence of co-ethnic communities, as well as asense of welcome from the population in general.84My sister is in Montreal right now for school and she says she has just recently found thefirst… person from Sudan in Montreal, and that’s amazing… Every street I turn I see a…person from Sudan. I really like that, even though I’m in a different place, I still get to seepeople from the same country… I really like Burnaby because there is a lot of diversity.My mom, she’s had a lot of help from people in Burnaby and locally, with finding jobsand just helping out the family. I have 7 siblings with a single mom.I like the people here. They are very welcoming to us. They show respect. When I was avolunteer, I like they smile at me every time. The people I like.Here I found it not too bad to settle in because there’s a lot of community things that youcan go to, to meet people. The schools are quite good for getting families together. Ifound it not too hard to start making a network. Obviously the first few months are hardanywhere, but I didn’t find it too bad here.When my son started kindergarten… they were very welcoming, saying, “Comevolunteer.” I volunteered a few times here and the teacher was very welcoming.Here people seem more open to taking in new people.An additional benefit of the walkability of the neighbourhoods was that public spaces wereoccupied by local residents at all times of the day. This created a lively street environment thatwas important to the neighbourhoods’ character.Participant A: [In other places of the region] you don’t find what you find here. Over inNorth Burnaby you have bigger houses and nice streets maybe, but there isno life in there. No motion, no movement. Here you feel…Participant B: Action! [Interrupting  P A and drawing laughter and enthusiasticagreement from all other participants.]Participant C: I like that the area has action.85P A: You can see motion, people moving around.The combination of a walkable, attractive built environment, and a community in whichparticipants felt welcomed meant that the neighbourhoods of Maywood and Richmond Park weremuch more desirable than other parts of the Vancouver region in which some participants hadlived.When I lived in North Burnaby and Coquitlam I had to ride to go to the library, I had toride to go to swimming pool, everywhere I had to ride. But in this here, very convenient.I walk!It’s not like Surrey. If you go to Surrey you see the places, like some villages… so darkand some places are in the bush… you cannot see the shopping centres [or] stores around.Everything looks like a rural area. But here you are just like in the city… the place isbeautiful, you can see the people around. Everything is good. The shopping centres, thecommunity centres, the schools, the environment… everything is there. If you want to seethe people you can see them. But if you are in Surrey you cannot see people around. Likejust walking… you have some people that you know, you can just see them anytime youwant. You are going out you can see them, say, “Hi,” have time, like two, three minutetalking to each other.At first it was after we just moved to Canada… we couldn’t meet so many people fromour community, it was far, so that’s why we moved [to Burnaby from Surrey].The Skytrain was clearly important to all participants, and the central location in the regionmeant that less time was spent commuting long distances. In many cases participants mentionedthat they or a family member could not drive, or that they did not own a car, so access to theSkytrain gave them a mobility that they would not have otherwise. Those participants or their86family members that did have access to a vehicle would often choose the SkyTrain over their carto get around.[T]he SkyTrain transportation is also good. With little kids and family it is veryconvenient. It’s a central place right, Burnaby is the central place I think. We moved toLangley but it was too hard to come to downtown. Burnaby is quite good. Everything isconvenient nearby.For us it’s the SkyTrain, because we don’t have a car so SkyTrain is very important forus. It’s a pretty central point. You can get downtown easily but it’s not the cost ofdowntown.We like very much this area because we live very close to SkyTrain station. It’s veryaccessible for my mother in law. We live together and she doesn’t drive and it’s veryaccessible for her.My husband works in (Metrotown) and at first we lived in Port Coquitlam, and he wasspending lots of time and money commuting, so we decided to move.I think it’s a little expensive, but at the same time it’s very convenient you don’t spend somuch time in the commute.I have a car, but if I have the opportunity to go somewhere on the SkyTrain, I prefer theSkyTrain.If you don’t have car you can just go shop around, you don’t have to go far. If you wantto go to Metrotown the transport is available. It doesn’t matter if you drive or you don’tdrive.Participants’ individual use of SkyTrain varied, but having access to rapid transit wasappreciated by all.CEJ: Does everyone use the SkyTrain?87Participant: Yes.CEJ: Would that be every day, once a week, once a month?Participant: Some of us daily, some of us whenever we need it.CEJ: So it’s a good thing to have in the neighbourhood? [This question drew anenthusiastic “Yes!” from all participants.]On aspect of the neighbourhoods that was not regarded as positive was the high cost of housingrelative to incomes.So many good things except rent is a little bit expensive.It is a bit expensive… we have everything around the neighbourhood, but the rent isexpensive.Also of concern to all participants was the expectation that rents would continue to increase overtime.CEJ: Do you expect that rents will increase in this area?Participant: It already is.Concerns over increasing rents translated into negative thoughts about having to live in otherparts of the Vancouver region. Some feared the difficulties they would face in finding affordablehousing of decent quality in a neighbourhood that would offer the benefits found in Burnaby.I was thinking about moving away from here and if I look at some other places, I don’tfeel comfortable moving out from Burnaby, it’s like I’m born in Burnaby… it’s thebackground, where I came from.88When I came here we were renting two bedrooms for $750 and it was a nice place. Nowit’s a lot higher than before and my children are growing. We need to move and I don’tknow where to go. I can’t afford to get out of that place because once I get out of thatplace I need to get a different place and it’s for sure going to be more expensive. And Idon’t want to go away from Burnaby because I like the school and I like everything.I pay $1005 in rent. If they increase that by anything, I will have to move. Maybe toCoquitlam, and over there is going to be like death, not life.They will push us to live in places like Surrey because it will be cheaper there forsomeone like me. It will be very difficult.5.3.4 Perception of Development in Richmond ParkThere was a significant difference between the neighbourhoods in how participants perceivedrecent development. In Richmond Park there have been a handful of developments ofcondominium towers in the last few years. These towers have been accompanied byimprovements to the public realm, such as a shopping centre, a new library and communitycentre. Participants in Richmond Park noticed and appreciated the positive effect that that newamenities were having on the area.We have the new library now, which is a good thing.We used to go to Metrotown to use the community centre, but now it is very close to ourneighbourhood so we can use it whenever we like. Especially our children, we can walkthere.Before the community centre, we used to be discouraged by the distance because to gothere and come back takes a while, but now it’s walking distance so we are using it asmuch as we can.89New amenities for the neighbourhood, made possible through density bonuses fromcondominium tower developments were seen as obviously good for residents.Participant: We have lived here for long, so it’s improving every time there is achange.CEJ: Do you like the changes?Participant: Yes, who doesn’t like good thing? [This comment drew laughter from theother participants.]Opinions on the towers themselves were somewhat ambivalent. Participants appreciated theaesthetic value of the towers and hoped that increased residential density in the area could makethe cost of housing more affordable.The towers are nice.A few of them are OK, but if we keep having more and more it’s going to be likedowntown…A few for aesthetics, it’s OK.At first the impression is that increasing high rises will lower rent.Right now we don’t have the intention to look for new house, but if they built moremaybe the price would be cheaper.An increase in residential density was welcomed by participants if it offered an opportunity tofind housing for their community, friends and children.CEJ: How would you feel about an increase in density in this neighbourhood?More residential towers?90Participant A: If the price is reasonable, we would like more houses.Participant B: It’s better to have more houses here.Participant C: Whether they build more it doesn’t matter as long as the price is high. Forus if the price is [not] going to be the same it makes no difference for us,whether they build more or not.Participant D: When I came here 2 years before [we found housing in this area], we askthem [if we could] rent house, but all of the apartment it was full. So Ihave to go to another area. So it’s better to have [more housing] becausemany people are coming, our children also [are getting older], so theyhave to [find a place of their own]. So it’s better for the future.CEJ: Are there lots of people in your community that want to live here butaren’t able?P B: Yes. We have so many friends; they ask us if there are new housesbecause they want to move to this area.However, there was a disconnect between the hope that an increase in residential density wouldimprove housing options and the reality that none of the participants lived in one of the newtowers, nor did they expect to in the future.CEJ: Does anyone live in one of the new towers? [All participants answeredno.]Participant: No. It’s expensive. We just think it’s expensive [so we don’t try to livethere].The perception that units in the new towers were too expensive was accompanied by a concernthat the improved amenities that accompanied new developments could drive up rents.91These [new developments in the area] increase the rent.With every improvement in the area, the rents increase. Those who suffer the most are us,people with limited income.5.3.5 Perception of Development in MaywoodIn Maywood the current development of towers was unanimously perceived as negative. Oneparticipant labeled development in the neighbourhood as an “attack of high rises,” and the towerswere viewed as an immediate threat. Participants were concerned that landlords were seeing thenew buildings as a future opportunity to sell their land to a developer, and that this was acting asa disincentive for landlords to properly maintain their buildings.I’ve noticed in just the two years I’ve been here I saw high rises over there, over there,over there [pointing south, west and north]. And I think they are very expensive. I’venoticed where we are staying, they are not changing and they are not renovating anymore.If it is old they are not changing, that is why we are looking [to move].[T]he new builds are so expensive to rent; there’s either old places that need renovatingthat they’re not doing because they’re waiting to just sell and then something new will goup that most people are priced out of it.Participants in Maywood also expressed feelings of exclusion. One form of the exclusion was aninability to influence the political process which approved the development of towers leadingparticipants to conclude that they had no control over the changes happening in theirneighbourhood.I think, for example the high rises in Metrotown area might be part of the plan of the Cityof Burnaby, because when they started, there is a new high rise coming out… because welive nearby that place we went to City Hall to have a public hearing. They invited people92from the neighbourhood and with the City Council; Mayor and everybody. There weresome people from the low-rise in that area, they talked and they said with all the newhigh rises coming out, that will push away low income families because they can’t affordthe high prices, but they like the convenience of living in the Metrotown area; theSkyTrain, library, shopping mall, everything. But the Mayor was there, the Councillorswere there, I think they had a public hearing but after that meeting, then I saw the signcame up already. So people can make their noise and go talk to the Mayor or something,but I don’t know if it’s going to change the plan of the city or not.There is so many high rises coming up, one after another, I think we can talk to the mediaor to the City Hall, but I don’t know how it’s going to affect the decisions of the City.The plan is changing the neighbourhood. What I see is just they have to build more highrises, they not care about the people living in this area… It is plan to build high rises andthe people’s concern is not their concern, because they can make more money frombuilding more high rises.If we could ask the Mayor to put up the high rise, but affordable. But who can talk tothem?We cannot stop this trend of having more high rises.The physical spaces of the towers themselves were perceived as another form of exclusion,because it was thought that the new units would be too expensive for participants to rent.Participant A: Everywhere you look it’s high rise, high rise and noise. Even my daughter,she is 8 years, youngest one, and we came out after swimming classes and[she said], “OK Mom, we can buy one of these beautiful [condos],” and Isaid, “No.” She thinks it cheap.Participant B: Yeah the same thing, my daughter she is just 11 and she says, “Mom, wecan take one condo. OK we will take?” Because in my family we have 493members and my husband is the only one who is having the income [wecould not afford to rent a condo].Many participants in Maywood feared that they would soon be pushed out of the neighbourhoodbecause it was changing so rapidly.This neighbourhood is changing fast… The construction is growing fast; it is changingthe landscape of the city. It is attack of high-rises. People who are interested indevelopment are coming to this neighbourhood and I know our neighbourhood ischanging so fast. 5 years ago I was here, living here. It was quiet, we never thought herewould change. And now we see every building in this area is going to be demolished [thiscomment drew general agreement from the other participants], it will be high rise, andpeople will change… I believe in next 5 years it will change and people will move fromhere. And it is so sad that change is going so fast… It is just so sad and what we see isdestruction, inconvenience… In this area all low rise buildings completely gone in 3years.I feel less secure [in tenure] than before... In my opinion they will destroy our oldbuilding because both sides they built a high rise. That means in maybe two years they[will replace our building with a high rise]. We have to think about the future. Move outor buy something. We are scared to move out…Especially on the south side of the Skytrain there, where they’re knocking down the lowrises and building the high rises, there’s going to be a completely different demographicof people moving into those high rises compared to the people who were living therebefore… And I think that the people who could have afforded to live there are going tobe shifted further along because they can’t afford to live there anymore.945.4 Understanding the Different Perceptions of DevelopmentI set out to understand why the perception of development would differ so much between the twoneighbourhoods and discovered a 2011 text amendment to Burnaby’s zoning bylaw. The textamendment introduced ‘s’ zoning, which applies a special status to all multiple-unit dwellingswithin Burnaby’s four town centres, and enrolls them into a density bonus program whichestablishes a negotiation framework for discretionary zoning between developers and the City ifBurnaby (Burnaby 2011). Developers are able to negotiate for an increase in the maximum floorarea ratio (FAR) of a site in exchange for the conservation or provision of amenities or non-market housing (Burnaby 2014).The ‘s’ zoning was originally intended to be implemented through individual Town Centredevelopment plan updates, but because there were numerous opportunities for the application ofthe ‘s’ category, the City of Burnaby considered it appropriate to fast-track the ‘s’ zoning byapproving it as a text amendment to the existing zoning bylaw (Burnaby 2011). A public hearingat Burnaby City Hall regarding ‘s’ zoning was held on 23 November 2010 and no submissionswere received from the public  regarding the text amendment (Burnaby 2010b). Publicconsultation regarding the text amendment was terminated at the end of the hearing and as of thispoint no further presentations regarding the text amendment would be received by Council(Burnaby 2010a). In each of Burnaby’s four Town Centres, the ‘s’ zoning category is applied toall multiple-unit residential buildings which are zoned as RM3, RM4, or RM5. Table 6 below,outlines some of the limits and requirements of these zoning designations.95Table 6 ‘s’ Zoning and Maximum FARThe FAR is a measure of the maximum buildable area of a development. Using RM4 as anexample; on a 1,000m² site with a maximum FAR of 2.0, a developer has a limit of up to2,000m² of buildable space. As 25% of the lot area may be covered, the structure that would takefull advantage of the maximum FAR would be an 8 storey building of 250m² per storey. In anRM4‘s’ zone the FAR could be increased to 3.6, the building that would take full advantage ofthis maximum FAR would be 15 storeys tall. As we can see in the table above, the application of‘s’ zoning has a remarkable potential to increase the maximum allowable FAR on a site andtherefore a remarkable increase in buildable area – and site value.In order for the ‘s’ zoning to be applied to a site it has to meet certain criteria. It must:1. be located in a town centre.2. be approved for density bonus in the community plan.3. be rezoned to a comprehensive development district.4. include in the comprehensive development plan the conservation or provision ofamenities or non-market housing.The conservation or provision of amenities or housing must be “equivalent in value to theincrease in the value of the lot attributable to the increase in floor area ratio” (Burnaby 2014).When the City of Burnaby approves the rezoning of a site to comprehensive development districtunder the ‘s’ zoning designation, the increase of maximum FAR increases the maximumbuildable area, which increases the value of the land itself (an effect known as ‘land lift’). It isZoning Category Max. Height Min. Storeys Max. Storeys Max. Lot Coverage Max. FAR Max. FAR 's'RM3 30m NA 3 NA 1.1 1.5RM4 30m 4 NA 25% 2.0 3.6RM5 55m 4 NA 30% 2.8 5.0Source: City of Burnaby Zoning Bylaw96the policy of ‘s’ zoning that the City of Burnaby capture the full value of the land lift from adeveloper through an investment in amenities or non-market housing on site, or as cash-in-lieu ofa physical amenity (Burnaby 2014). Figure 17 below, shows that after the introduction of ‘s’zoning in 2011, the number of permits granted by the City of Burnaby for the demolition ofapartment units has increased sharply .Figure 17 Apartment Demolition Permits in Burnaby, 2002-2014This uptick in demolitions is due to the fact that ‘s’ zoning includes almost all of the low-risePBRH in Burnaby under the density bonus program. In 2012, there were 363 PBRH properties inBurnaby that were built before 1980, and almost all of this housing stock was located in areas ofMaywood and Richmond Park which lie within the boundary of Metrotown and Edmonds TownCentre respectively (Coriolis 2012a, 2012b). Although detailed information on the location ofapartments demolished in 2012 was not readily available, the vast majority of apartmentdemolition permits approved in 2013 and 2014 were located in Metrotown, within or just outsidethe boundary of Maywood. The geographical concentration of apartment demolition permits in97and around Maywood might explain why participants in Maywood had such a differentperception of development from those in Richmond Park.One caveat of figure 17 is that a fire caused the demolition of one building containing 34apartment units in 2013. The remaining 77 demolished apartment units were located at 6350-6550 Nelson Avenue in Metrotown, just east of Maywood. The estimated value of the amenitybonus for this project was $9.5 million, given to the City of Burnaby by a developer as cash inlieu of a physical amenity on site. At the public hearing which approved the rezoning of this site,concerns over the loss of affordable housing in the neighbourhood were raised. The City’sresponse was this:It is acknowledged that the issue of housing affordability is complex and challenging andinfluenced by many external factors such as market conditions – supply and demand,projected population growth, income, and market costs of land and building construction.While these existing units may provide a measure of affordable housing within the TownCentre, like many buildings that are nearing the end of their life-cycle, they are advancedfor redevelopment based on market conditions, and as it becomes increasinglyuneconomic to continue to repair and maintain older building stock as they age (Burnaby2013a).A critique of this line of reasoning is that it applies to an entire region of affordable housing.Under this logic, almost all of the aging stock of PBRH in Burnaby is a candidate forredevelopment. A 2012 report for Metro Vancouver concluded that about 3% of Burnaby’spurpose-built rental housing stock was at risk of redevelopment (Coriolis 2012b). The reportcautioned that this figure was likely to rise in the future as the value of land often increases fasterthan the value of apartment buildings. Had this report taken into account  the introduction of ‘s’98zoning, which has the potential to drastically increase the value of a parcel of land, the estimateof risk would surely have been much higher.These concerns were raised in an interview with an elected government official. The official wasaware that the housing stock of Maywood and Edmonds was unique in the Vancouver regionbecause there are affordable rental apartments which are larger than most, making it attractive tolow-income families. Of concern to the City of Burnaby was that the majority of the PBRH hadbeen built in the 1960s and 1970s and several apartment buildings were deteriorating as theyapproached the end of their useful life. This housing had not been designed to exist for more than50 or 60 years and none of the buildings had been maintained to extend their life cycle beyondthat. Speculation on the value of land and the expectation of increased density in the area actedas a disincentive for landlords to properly maintain the quality of their buildings. As the housingstock depreciated it became affordable for low-income people, but in some cases the housingdeteriorated to a point at which the City was forced to intervene to address poor livingconditions. The City became concerned that the deteriorating housing stock could contribute to afurther concentration of poverty in these areas which could lead to social issues that the Citywould not be equipped to manage. Ideally the City would have liked to introduce a diversity ofhousing within these areas to achieve a ‘social mix’ of income levels. However, without supportfrom the federal and provincial levels of government to provide incentives or funds for social orrental housing, it was seen as inevitable that densification would lead to gentrification anddisplacement.  The City of Burnaby approached the Provincial Government with a plan to createa zone exclusively for rental housing; a plan which was rejected by the Province. Efforts toinvest municipal funds in subsidized housing were unable to have a significant impact upon the99demand for affordable housing and investment of funds directly into housing reduced the City’sability to fulfil other responsibilities. It was determined that funds gained through densitybonuses could be better directed at the provision of amenities or put into an affordable housingfund which could be used to offset permitting and licensing costs for projects that sought todevelop subsidized housing. The decision to introduce ‘s’ zoning came after a long and arduousprocess, during which many solutions were considered, but none were found to be ideal. In theend it was decided that redevelopment in Maywood would be facilitated through ‘s’ zoning.You’ve got to remember we've been resisting for 25 years the redevelopment of this area,and there’s only so far you can go before the writing is on the wall.There's no easy equation and I can tell you, we went through a study early on in regard toMetrotown, looking at ways we could try to stimulate the recreation of that amount of[rental] housing that existed, by the new development density that we brought in.Impossible. Every consultant we had said, “Can’t do it, can’t do it, can’t do it.” Thenumbers don’t work. And you can’t get that rental housing built because rental housing isjust not marketable…So what I'm doing is accepting the inevitable, which is that you can’t have this low adensity around a Skytrain station in the middle of an urban centre. And so I've only gotthat broad sort of brush to paint with. Which is saying by creating more housing, thenhopefully, on the trickle-down theory, is that eventually this is going to create betterhousing opportunities for others. I can’t hold back the sea of change. We eventually,despite being a left wing council, one that’s very conscious of these issues, we came tothe conclusion that we couldn’t continue to stop development from happening in theMaywood area [just] because we thought that the people in there should somehow beprotected from that reality. Because we were no longer doing them any favours as therewas such a downturn in the quality of housing that they were turning into fire traps, rattraps and it was not going well. We spent a lot of money on consultants and trying to find100some magic solution that didn’t exist. We worked very hard at it and no matter how manymeetings we had, and no matter what we went through, we couldn’t find a way. Even ifwe took all of what we had and increased density in ‘s’ zoning and threw [all of the fundscollected through density bonuses] back in, we still couldn’t accomplish that goal [ofretaining the stock of rental housing]. At that point people threw up their hands and said,“We’ve just got to allow this…” Unless we were told that we could [zone the areaexclusively for] rental housing, and we could have froze everything and said, “You're allin a rental housing character.” Then that would be a successful way to move on. Asidefrom that, as soon as we got to the point that we were allowing the first high-rise in, weknew that the flood gates would open. Either you freeze this area and you ghettoize it,which was what it was becoming, or you say, “Alright, once I open this door we knowthat everybody’s going to be saying, ‘What about me? You let them, what about myproperty?’” And it’s hard to deny that in Metrotown, which is a regional town centre andthe area that’s closest to transit with the highest transit ridership station in the LowerMainland that we’re not going to allow density... That would work against everythingwe're saying in every other area. It was never a comfortable and easy decision and I thinkyou should be aware of that. We don’t make these decisions in a cavalier way. Weagonize over the decisions we’re making.I can’t stimulate the other orders of government… to be involved in any of this. Andwithout their participation, I've got no option to be able to deal with this [housing issue].After several attempts to find another solution to the issue of deteriorating housing, the City ofBurnaby decided that the redevelopment of Maywood would take place, and the demolition ofPBRH in this neighbourhood would be permitted. It is obvious that the City of Burnaby founditself in a difficult position with hard choices to make. In the face of deteriorating housingquality and a vacuum of support from other levels of government the decision was made tofacilitate redevelopment in Maywood through the introduction of ‘s’ zoning, and the inevitabledemolition of deteriorating PBRH had to be accepted. However, there are unintended101consequences to this line of reasoning. Now that the City of Burnaby has introduced ‘s’ zoning,low-rise PBRH in Maywood that is in good condition is as much a candidate for redevelopmentas low-rise rental housing in poor condition. As Coriolis Consulting found, age and condition areonly part of a building’s redevelopment potential. They note that, “The ratio of existingfloorspace to permitted floorspace is a good indicator of whether a property is financiallyattractive for redevelopment” (2012a). The introduction of ‘s’ zoning has skewed that ratio inMaywood and facilitated the demolition of 63 units of PBRH at 6225 and 6255 Cassie Avenue,which according to a council report was in ‘fine condition’ when the site containing a low-risePBRH building was rezoned to RM5s with a maximum FAR of 6.28 (Burnaby 2011, 2013b).The broad brush of ‘s’ zoning addresses the issue of deteriorating housing, but it also colourshousing in good condition green for redevelopment.5.5 ConclusionMy attention was brought to Maywood and Richmond Park by the indicators of low income,rental housing vulnerability, recent immigration, and refugee location. Community profiles basedon 2006 census data for Maywood and Richmond Park showed that these neighbourhoods areoccupied mostly by immigrants and that between 1986 and 2006 the percentage of the populationthat spoke English at home dropped drastically, reflecting changing trends in country of originfor immigrants to Canada, and suggesting that the share of visible minorities in theseneighbourhoods increased sharply. These two neighbourhoods are predominantly occupied byrenters in apartments and the majority of these renters were spending more than 30% of theirmonthly income on housing. The prevalence of low income families in Maywood and RichmondPark was much higher than for Burnaby as a whole. Based on this evidence I expected to find102neighbourhoods in decline, but in the process of conducting focus groups in Maywood andRichmond Park, I found vibrant communities that were appreciated by local residents.The neighbourhoods offer a variety of benefits to the participants of my focus groups; all ofwhom were newcomers to Canada, many self-identifying as a visible minority, and almostentirely renters who struggled with the high cost of rent relative to their household incomes.These neighbourhoods are walkable, well-served by rapid transit, central to the region, and haveeasy access to valued services and public amenities. They are more than districts of low-incomerenters living in poor housing; they are desirable, complete, welcoming communities that offer aquality of life that is difficult for low income people to find in other parts of the region,particularly if they are refugees or newcomers to Canada.The high cost of housing relative to incomes, and the steady increase of rents over time put thoseon limited and fixed incomes in housing stress and the possibility of having to move to findlower cost housing was of great concern. There has been very little in the way of affordablehousing created in the Vancouver region in the past 20 years, and rental housing supply hasfailed to keep up with demand (Hiebert et al. 2006). Therefore, it is highly unlikely that theoccupants of the 200-plus apartment units for which demolition permits were issued in the last 3years would have been able to find affordable housing in their neighbourhood, and an electedgovernment official agreed with this assessment.Density bonuses gained through ‘s’ zoning have made it possible to bring new amenities andservices to Maywood and Richmond Park, and the residents of these neighbourhoods havebenefitted as a result. Redevelopment through ‘s’ zoning is continuing in these neighbourhoods,but the experience of redevelopment differs greatly between the two. The perception of103development in Richmond Park was somewhat ambivalent as the construction of condominiumtowers had so far taken place without the demolition of low-rise PBRH buildings. Recentimprovements in the neighbourhood such as a new library and community recreation centre weremade possible through the density bonus program for new developments. The spatialconcentration of demolition permits within or close to the boundary of Maywood meant thatredevelopment there was viewed with much greater concern.  In Maywood several low-risePBRH buildings have been demolished to make way for towers, and participants in thisneighbourhood expressed feelings of helplessness to intervene in the political process that wasfacilitating redevelopment.I must admit that I thought it was an exaggeration when one of the focus group participants saidthat in three years all of the low-rise buildings in Maywood would be demolished, but afterinterviewing an elected government official who was involved to the introduction of ‘s’ zoning,I’ve come to question how far from the truth this statement might be. The process through whichthe redevelopment of Maywood will take place will be mediated through the market, and theneighbourhood’s transformation will be influenced by “external factors such as marketconditions” and the probability that a given building will be “advanced for redevelopment” willdepend upon “market conditions,” but as the case of 6225-6255 Cassie Ave. shows, this will notsolely depend upon whether buildings are “nearing the end of their life-cycle” (Burnaby 2013a).Although it is unlikely that the entire stock of low-rise PBRH in Maywood will be demolished inthree years, it is likely to be just a matter of time. The “writing is on the wall” that the days oflow-rise purpose-built rental apartment buildings in Maywood are numbered. Whether it takesthree years or twenty, the floodgates have opened.104It is understandable how the City of Burnaby arrived at the decision to facilitate redevelopmentin Maywood through ‘s’ zoning. The municipality is operating in a context of expandingresponsibilities and limited resources, with little support from the provincial or federal levels ofgovernment to tackle issues related to housing (Hiebert et al. 2006). However, in attempting toaddress a concern that a deteriorating housing geography would lead to a deteriorating socialgeography, the blunt tool of ‘s’ zoning made the redevelopment of all low-rise PBRH inMaywood financially feasible, regardless of the actual condition of the building. Even buildingsthat are in ‘fair’ condition are now attractive for redevelopment because the ratio of permittedfloorspace to existing floorspace can be ‘s’ zoned to the point where the value of the landbecomes much greater than the value of the building. Under these conditions age and state ofrepair become irrelevant, which ironically was the reason why ‘s’ zoning was introduced in thefirst place.105Chapter 6: ConclusionThe main objectives of this thesis were:1. To identify which selected characteristics of neighbourhoods are associated withdifferences in the average incomes of Greater Vancouver CTs and changes in the averageincome of CTs over time.2. To discover what may have caused the emergence of a district of low- and very low-income CTs along the SkyTrain Expo Line from East Vancouver to North Surrey.3. To investigate how regional processes of socio-spatial polarization are expressed inparticular neighbourhoods.Method and data are extremely important for research in this area, as research design decisionscan have an influence on findings (see Hamnett 2003, 2009; Watt 2008; Hamnett & Butler 2008,2013; Davidson & Wyly 2012). With this in mind, I adopted a mixed methods approach toconduct research at three scales of analysis. This approach allowed me to use quantitativeanalysis to establish a regional context of income inequality that guided qualitative research atthe sub-regional and neighbourhood scales.In chapter 3 extensive methods were employed in order to gain insight into processes of incomeinequality and change over time in the Vancouver CMA. Although the results must beinterpreted with a measure of caution and several of the key findings must be seen asassociations with, not causes of, income polarization, the results of this chapter are useful forguiding thinking about income polarization at the regional scale. Of the many findings in chapter3, those that have the most relevance to the major themes of this thesis are the significantassociations between the variance in average CT incomes and changes in average incomes, withvisible minorities and recent immigrants who engage in residential crowding in poor quality106apartments as an affordability strategy. The interpretation of these key findings is generallysupportive of the global cities hypothesis, in which the integration of the local urban economyinto global flows of capital and people will shape the polarization of the labour market, leadingto increased income inequality. Immigration is an important aspect of this process as it is oftennew immigrants that find work in the low-wage service sector (Friedman 1986; Sassen 1988).This is not to deny the merit of the SBTC thesis. The labour market of the City of Vancouver hastransitioned from a resource extraction and manufacturing base, and witnessed the rise of a post-industrial service economy (Miro 2011). In this context it follows that the relative demand forjob skills would change, but the findings of this thesis place immigration at the centre of risingincome inequality, which is not consistent with the SBTC thesis. The distinctive effects ofneoliberal policies were also evident in the helplessness of a municipal government to protectaffordable housing as the existing stock aged, in the absence of any political will from seniorlevels of government to use revenue sources to build new subsidised housing. In true neoliberalfashion the state withdrew from social need opening the door to private market redevelopment.The major findings of chapter 3 were generally supported by interviews with key informants inchapter 4. One of the expressions of increasing income inequality in the Vancouver CMA hasbeen the emergence of a low income corridor which follows the SkyTrain Expo Line. Thiscorridor follows the region’s oldest overland migration route and a former rail right-of-way,which for one key informant explains lower average CT incomes because property values tend tobe lower near rail corridors.  As this was a rail and automobile corridor long before rapid transitwas introduced, and the concentration of low-rise PBRH pre-dates the SkyTrain Line, rapidtransit could not been seen an original of low average income CTs in the corridor. Over the107course of conducting semi-structured interviews with 14 key informants, several major themesemerged concerning low-income census tracts in the SkyTrain corridor:1. A basic factor is the significant presence of recent immigrants and refugees in thecorridor, with visible minority status being an important consideration.2. Housing is central to the emergence of the low-income corridor. All key informants citedthe importance of rental housing for explaining the presence of low incomes in thecorridor, and although rental housing was not found to be statistically significant inchapter 3, the significance of apartments is supported by statistical analysis. In particular,it was low-rise, wood-frame, PBRH apartment buildings that were most often cited bykey informants as explaining the presence of particularly low incomes in the corridor.3. A major finding which emerged through interviews with key informants was theimportance of contingent high-density redevelopment in the corridor. Whereas thedevelopment of new dwellings was not statistically significant in chapter 3, in theSkyTrain corridor high-density redevelopment near stations emerged as a major theme.While the expression of development and the community’s response varied across themunicipalities that the corridor runs through, a common justification for increasingdensity was the need to accommodate population growth in an environmentallysustainable way by concentrating high-density development near rapid transit stations.Whether it is through the politically mediated routing of transportation corridors or accidents ofplanning, the SkyTrain corridor is home to a district of affordable rental apartments and suiteswhich provide opportunities for recent immigrants and refugees, many of whom are of visibleminority status, to find housing that is large enough for families. That the average income of CTsshould be lower in this corridor is at least partly a result of this housing stock and its occupancyby recent immigrants, refugees and other vulnerable groups. Housing emerged as a central factorin all three scales of analysis and I argue that housing has not received adequate attention inresearch on urban income inequality and socio-spatial polarization. Hiebert (2009) illustrates the108polarized housing outcomes among immigrants to Canada, and as incomes continue to becomemore unequal, housing will continue to play a key role in the neighbourhood channelling of low-income people.As the population of the region is expected to continue to grow in the coming decades, municipalgovernments are looking for ways to accommodate this growth in an environmentally sustainableway. For many, adding density at places where there is good access to public transit makes sensefrom a planning perspective. In those places well-served by rapid transit, where municipalgovernments introduce policies to facilitate increased densities in order to upgrade the physicaland social geography of an area, and where the private sector can make a profit by capitalizingon a value gap, densification is likely to happen. In some parts of the region, increased densitiesare opposed by the community, in others there is ambivalence, and in some there is support.There are issues associated with contingent development in the corridor. The most prominentissues were speculation and disinvestment by land and property owners. As land owners areaware of policies which encourage high-density development in the corridor, there is adisincentive to invest in the maintenance of properties as the potential value of land afterrezoning far outstrips the value of land in its current use. Speculation can cause issues such asvacant lots in key locations, rented single-family detached houses used for illegal activities, andrental apartment buildings falling into disrepair as landlords weigh the cost of maintenanceagainst the potential for sale.Although the causes are difficult to identify, the concentration of affordable, aging, low-risePBRH apartments found along the SkyTrain Line corridor has become an increasingly rarehousing type which offers opportunities to those near the bottom of the region’s housing market.109However, as densification near rapid transit has become a common policy framework toaccommodate population growth, a conflict has emerged between the preservation of affordablerental housing and densification near SkyTrain stations.All of the major findings and significant themes that emerged in chapters 3 and 4 are relevant tothe neighbourhoods selected for further interrogation in chapter 5. The neighbourhoods ofMaywood and Richmond Park in Burnaby exhibit several of the key characteristics identified atboth the regional and sub-regional scales. Many of the residents are recent immigrants of visibleminority status, and a number are refugees. Half of all dwellings in 2006 were low-rise, wood-frame, PBRH apartments. High density development is also present in these neighbourhoods,with condominium towers perceived to be a major threat to security of tenure in Maywood inparticular.Contradictions and issues arise when areas near transit that are in decline from one perspective orscale of analysis, are seen by local residents as vibrant communities providing a high quality oflife to those with limited incomes. This is a contradiction that speaks to a larger methodologicalmatter. It is unlikely that quantitative analysis at this point in time would have identifiedMaywood as a site of gentrification. Regional indicators suggest that this neighbourhood is oneof the most disadvantaged, but from a local perspective, gentrification is clearly taking place andis likely to continue. I must admit that I experienced this contradiction first-hand. Whereas theregional indicators suggested that the neighbourhoods selected would be problematic, and it wasclearly the impression of one elected government official that they were problematicneighbourhoods on their way to becoming ghettos, my assumptions about these neighbourhoodswere challenged by the unanimous feedback from focus group participants. I cannot claim that110participants were representative of their neighbourhood as there was a selection bias in that eachindividual had to have prior contact with either South Burnaby Neighbourhood House or theMOSAIC BC office in Edmonds; but according to them, these are vibrant, liveableneighbourhoods that offer accessibility, social services, public amenities, communityengagement and co-ethnic networks of support.Regardless of the obvious benefits that Maywood offers to low-income residents, it is difficult toargue that affordable housing should be protected if it is nearing the end of its life cycle, ispoorly maintained, is becoming a liability for the municipality, and is in a location in which thevalue of land is far greater than that of the building on site. According to a “planningperspective” and market logic, demolition is inevitable. However, it has been recognized thatsince the 1980s the provision of affordable and rental housing in the Vancouver CMA has notkept up with demand (Hiebert et al. 2006). When these affordable rental apartments inMaywood, built between 1950 and the early 1980s are demolished, an irreplaceable element ofthe region’s housing stock will be lost. In addition to the loss of this increasingly rare type ofhousing, for the displaced residents of these buildings there are no other viable housing optionsin their neighbourhoods, forcing them to look for housing in more peripheral areas. Affordablehousing in the Vancouver CMA is a regional issue, and the tension between population growth,transit-oriented development, and gentrification is in need of further critical study (Kahn 2007;Duncan 2011; Grube-Cavers & Patterson 2014).A text amendment to Burnaby’s zoning bylaw has exposed to demolition a district of affordablerental housing in the Maywood neighbourhood of Burnaby. When the market deems, thishousing will be demolished. It is highly unlikely that current residents will be able to find111housing that meets their needs at a cost they can afford in the neighbourhood they have come tocall home. Displaced residents will have no alternative than to move further into the moreperipheral districts of the region. There are established co-ethnic communities in Maywood andRichmond Park that offer networks of support for newcomers to Canada. Should thesecommunities be disrupted, a valuable resource for newcomers could be lost.It is beyond the scope of this project to study specific processes of displacement and theirconsequences, but this is a project worthy of further study. I do not wish to make an alarmistclaim that all affordable housing in the corridor will be demolished within the next few years;this is a corridor of contingent development and the future of affordable housing in the corridoris far from predetermined. However, in those places where municipal policy and the marketagree that redevelopment should take place, affordable rental housing near SkyTrain stations issure to come under immense redevelopment pressure. As municipalities expect to see theirpopulations grow, they choose to add density near transit in order to reduce reliance uponautomobiles, but these policies could reduce access to transit for those who need it most: recentimmigrants, refugees and other vulnerable households on limited incomes without access to acar. Now that municipal policy and the market agree that redevelopment should happen atSkyTrain stations, those who live in aging affordable rental housing find themselves in the wayof change. It is unfortunate that in Maywood so many of these people are refugees andimmigrants who are struggling in the labour market. From an equity perspective it makes littlesense to provide housing near rapid transit for those with relatively less need for it. For low-income people and recent immigrants, having access to public transit is particularly importantand demolishing affordable housing near transit stations denies them access to this needed112amenity. As visible minorities, recent immigrants and refugees struggle in the Canadian labourmarket (Hiebert et al. 2006; Oreopoulos 2008; Hou & Coulombe 2010; Pendakur & Pendakur2011), having access to rapid transit for commutes to work is especially important, particularlywhen the work performed is not particularly well-paid. Although it is beyond the scope of thisproject to quantify the social costs of long work commutes, it follows that to deny low-incomeparents time with their children by displacing them further from their place of work could haveharmful effects, especially if this hampers the integration of the children of immigrants toCanada.The creation of a theory of urban income inequality is a work in progress. The world city andSBTC hypotheses are not conceptually robust enough to capture the myriad ways in which urbaninequality becomes manifest in a particular place. Processes of deindustrialization, diminishingunion density, immigration rates, levels of educational attainment and the policy emphases ofneoliberalism are all causes of income inequality, but as of yet no theoretical framework hassuccessfully accounted for all of these factors as they relate to globalization and economicrestructuring. A goal of geographers should be the creation of a theory of urban incomeinequality that accounts for the life chances of individuals in a given labour market as it relates tothe economic base of the city. The development of a standard measure of inequality, based onreliable and widely available data, which uses a well-recognized and replicable method, is animportant project to undertake in order to produce results which can be compared across contextsand time periods. This should be of primary concern to research on urban income inequality inCanada, as the scrapping of the mandatory long-form Canadian Census and its replacement bythe voluntary National Household Survey in 2011, has compromised the reliability of census-113derived income data (Hulchanski et al. 2013). With these considerations in mind, this thesis hasstarted from the methodology of Hulchanski et al. (2007) and Ley & Lynch (2012), and hasfurther interrogated the findings of Ley & Lynch (2012) at the regional, sub-regional andneighbourhood scale using qualitative sources as well as data bases. Geographers must beconcerned with urban income inequality as it relates to income segregation at the neighbourhoodscale, as it is especially at this scale that growing international income inequality and socio-spatial polarization hits the ground. This is particularly true in Canada as research onneighbourhood poverty and segregation is sparse (Chen et al. 2012) and Canadian cities arefinding themselves increasingly economically divided without knowing why (David Hulchanskiquoted in Paperny 2010). Fortunately there are some valuable precedents in this field (Ley &Smith 2000; Ades et al. 2012) and academic interest in neighbourhoods is building in Canada(Oreopoulos 2008).114BibliographyAdes, J., P. Apparicio and A. M. Seguin (2012) “Are new patterns of low-income distributionemerging in Canadian metroplolitan areas?” The Canadian Geographer. 56 (3): 339-361.Autor, D. H., L. F. Katz and M. Kearney (2006) “The Polarization of the U.S. Labor Market”American Economic Review 96 (2): 189-194.Autor, D. H., L. F. Katz and M. Kearney (2008) “Trends in U.S. wage inequality: revising therevisionists” The Review of Economics and Statistics 90 (2): 300-323.Beasley, L. B. (1976) A Design Probe Comparison of Regional and Municipal Attitudes TowardRegional Town Centres: Case Study in Burnaby, B.C. MA Thesis. The School ofCommunity and Regional Planning. The University of British Columbia.Beeson, P., L. Shore-Sheppard and T. Watson (2010) “Local Fiscal Policies and Urban WageStructures” Public Finance Review 38 (5): 540-584.Bolton K. C. and S. Breau (2012) “Growing unequal? Changes in the distribution of earningsacross Canadian cities” Urban Studies 49 (6): 1377–1396.Borel-Saladin, J., and O. Crankshaw (2009) “Social Polarisation or Professionalisation? AnotherLook at Theory and Evidence on Deindustrialisation and the Rise of the Service Sector”Urban Studies 46: 645-664.Bound, J. and G. Johnson (1992) “Changes in the structure of wages in the 1980s: an evaluationof alternative explanations” American Economic Review 82 (3): 37 1-392.Brunner, L. R. and C. Friesen (2011) “Changing faces, changing neighbourhoods: Government-assisted refugee settlement patterns in Metro Vancouver 2005––2009” Our Diverse Cities8 (Spring).Brzozowski, M., M. Gervais, P. Klein and M. Suzuki (2010) “Consumption, income, and wealthinequality in Canada” Review of Economic Dynamics 13 (1): 52-75.Burnaby, City of (1974) Rental Conversion Control Policy. Adopted by Burnaby City CouncilJuly 1974. Accessed online 6 June 2014 at:, City of (2010a) Public Hearing Agenda – 2010 November 23. Accessed online 18 July2014 at:, City of (2010b) Public Hearing – 2010 November 23.Accessed online 18 July 2014 at:, City of (2011) “Application of the Supplemental Community Benefit Density BonusProgram”. Community Development Committee Council Report. 7 March 2011.Burnaby, City of (2013a) “Rezoning Reference #12-15 6350 and 6550 Avenue Response toPublic Hearing Issues”. Council Report. 8 Jan 2013.Burnaby, City of (2013b) “Rezoning Reference #12-24 High Density High Rise Above aCommercial Office-Retail Podium”. Council Report. 6 May 2013.Burnaby, City of (2014) “Section 6.22 Density Bonus”. Burnaby Zoning Bylaw. 22 Jan 2014.Burnaby Newsleader (2014) “EDMONDS PULSE: Just watch Edmonds grow” BurnabyNewsleader. 28 March 2014.Butler, T., C. Hamnett and M. Ramsden (2008) “Inward and upward? Marking out social classchange in London 1981–2001” Urban Studies 45: 67–88.Carter, T. (1997) “Current practices for affordable housing: The Canadian context” HousingPolicy Debate 8 (3): 593-631.Cassiers, T. and C. Kesteloot (2012). “Socio-spatial Inequalities and Social Cohesion inEuroppean Cities” Urban Studies 49: 1909-1924.Chen, W., J. Myles, and G. Picot (2012). “Why Have Poorer Neighbourhoods StagnatedEconomically while the Richer Have Flourished? Neighbourhood Income Inequality inCanadian Cities” Urban Studies 49 (4): 877-896.Chrunik, E. (2007) “Kerrisdale Without the Canada Line” Vancouver Observer. 27 July 2007.Accessed online 14 July 2014 at:, Yolande (2013) “Critics decry Grandview-Woodland development plan” The GeorgiaStraight. 3 July 2013.Conference Board of Canada (2011) Hot Topic: Canada Inequality. Ottawa: Conference Boardof Canada. Accessed online 8 June 2014 at: Consulting  Corp. (2012a) Metro Vancouver Purpose-Built Rental Housing Inventoryand Risk Analysis. Report prepared for Metro Vancouver. May 2012.Coriolis Consulting  Corp. (2012b) Metro Vancouver Purpose-Built Rental Housing Inventoryand Risk Analysis: City of Burnaby Profile. Report prepared for Metro Vancouver. May2012.Coriolis Consulting Corp. (2013) Strengthening the Coordination between Land Use andTransportation. RTS Working Paper 3.100B, Draft Version 2.2. April 2013.116CTV News (2014) “B.C. tables legislation on transit referendum and TransLink governance”CTV News Vancouver. 27 March 2014. Accessed online 14 July 2014 at:, M. and L. Lees (2010) “New-Build Gentrification: Its Histories, Trajectories, andCritical Geographies” Population, Space and Place. 16: 395-411.Davidson, M. and E. Wyly (2012) “Class-ifying London: questions of social division and spaceclaims in the post-industrial city” City 16 (4): 395-421.Doussard, M., J. Peck and N. Theodore (2009). “After Deindustrialization: Uneven Growth andEconomic Inequality in ‘Postindustrial’ Chicago”. Economic Geography 85 (2): 185-207.Duncan, M. (2011) “The Impact of Transit-oriented Development on Housing Prices in SanDiego, CA” Urban Studies 48 (1): 101-127.Frenette, M., D. Green and K. Milligan (2007) “The tale of the tails: Canadian income inequalityin the 1980s and 1990s” Canadian Journal of Economics 40 (3): 735-764.Friedman, J. (1986) “The world city hypothesis” Development and Change 17 (1): 69–84.Friesen, C. and C. Daily (2014) Refugee Newcomers in Metro Vancouver: Changing Faces andNeighbourhoods 2010-2013. May 2014.Glaeser, E., M. E. Kahn and J. Rappaport (2008) “Why do the poor live in cities? The role ofpublic transportation” Journal of Urban Economics 63: 1-24.Glaeser, E., M. Resseger and K. Tobio (2009) “Inequality in Cities” Journal of Regional Science40 (4): 617-646.Green, D. A. and J. R. Kesselman (2006) Dimensions of Inequality in Canada. Vancouver:University of British Columbia PressGrube-Carvers, A. and Z. Patterson (2014) “Urban rapid rail transit and gentrification inCanadian urban centres: A survival analysis approach” Urban Studies. published online18 March 2014.Haan, M. (2010) “The Residential Crowding of Immigrants to Canada” Canadian Issues:Newcomers’ Experiences of Housing and Homelessness in Canada. Fall 2010.Hamnett, C. (1994) “Social polarisation in global cities: theory and evidence” Urban Studies 31(3): 401–425.Hamnett, C. (1996) “Social Polarisation, Economic Restructuring and Welfare State Regimes”Urban Studies 33: 1407-1430.117Hamnett, C. (2003) Unequal City: London in the Global Arena. London: Routledge.Hamnett, C. (2009) “The new Mikado? Tom Slater, gentrification, and displacement” City 13(4): 476–482.Hamnett, C. and T. Butler (2013) “Re-classifying London: a growing middle class and increasinginequality” City 17 (2): 197-208.Harvey, D. (2005) A brief history of neoliberalism. Oxford University Press, 2005.Heathcote, J., F. Perri and G. L. Violante (2010) “Unequal we stand: An empirical analysis ofeconomic inequality in the United States, 1967-2006” Review of Economic Dynamics 13:15-51Hiebert, D. (2006) “Winning, losing, and still playing the game: The political economy ofimmigration in Canada”. Tijdschrift voor economische en sociale geografie 97(1): 38-48.Hiebert, D. (2009) “Newcomers in the Canadian housing market: a longitudinal study, 2001-2005” The Canadian Geographer 53 (3): 269-287.Hiebert, D., P. Mendez and E. Wyly (2006) The Housing Situation and Needs of RecentImmigrants in the Vancouver CMA. Canada Mortgage and Housing CorporationResearch Report.Hiebert, D. and C. E. Jones (2014) Social Atlas of Immigrants in Canada’s largest CensusMetropolitan Areas. (forthcoming)Hou, F. and S. Coulombe (2010) “Earnings Gaps for Canadian-Born Visible Minorities in thePublic and Private Sectors” Canadian Public Policy 36 (1): 29-43.Hulchanski, D., et al. (2007) The Three Cities within Toronto. Centre for Urban & CommunityStudies, Univ. Toronto. Research Bulletin 41.Hulchanski, D., R. Murdie, A. Walks and L. Bourne (2013) “Canada’s voluntary census isworthless. Here’s why” The Globe and Mail. 4 Oct 2013.Human Resources and Social Development Canada (HRSDC) (2014) Indicators of Well-being inCanada. Accessed online 6 June 2014 at:, B. (2013) “Transit fuelling Burnaby’s growth” The Globe and Mail. 6 March 2013.Juhn, C., K. Murphy and B. Pierce (1993) “Wage inequality and the rise in returns to skill”Journal of Political Economics 101 (3): 4 10-442.118Kahn, M. E. (2007) “Gentrification Trends in New Transit-Oriented Communities: Evidencefrom 14 Cities that Expanded and Built Rail Transit Systems” Real Estate Economics 35(2): 155-182.Katz, L. and K. Murphy (1992) “Changes in relative wages, 1963-1987: supply and demandfactors” Quarterly Journal of Economics 107 (l): 35-78.Korpi, M. (2008) “Does size of local labour markets affect wage inequality? a rank-size rule ofincome distribution” Journal of Economic Geography 8: 211-237.Krueger, A. (1993) “How computers have changed the wage structure: evidence from microdata,1984- 1989” Quarterly Journal of Economics 108 (l): 33-60.Lemieux, T. (2008) “The Changing Nature of Wage Inequality” Journal of PopulationEconomics 21 (1): 21-48.Levy, F. and R. Murnane (1992) “U.S. earnings levels and earnings inequality: a review of recenttrends and proposed explanations” Journal of Economic Literature 30 (3): 1333-1381.Ley, D. (1999) “Myths and Meanings of Immigration and the Metropolis” The CanadianGeographer 43 (1): 2-19.Ley, D. and H. Smith (2000) “Relations between deprivation and immigrant groups in largeCanadian cities” Urban Studies 37: 37-62.Ley, D. and N. Lynch (2012) Divisions and Disparities in Lotus Land. Cities Centre, Univ.Toronto. Discussion Paper 223.Lee, J. (2012) “19-storey Mount Pleasant tower wins Vancouver council approval” TheVancouver Sun. 17 April 2012.Mendelson, M. and K. Battle (2011) Fixing the Hole in EI. Background paper, Mowatt Centrefor Policy Innovation. Univ. of Toronto.Metcalfe, B. (2013) “Density inevitable; the issue is planning” The Vancouver Sun. 5 Nov 2013.Metro Vancouver (2013) 2011 National Household Survey – Bulletin #7: Labour Force Activityin Metro Vancouver. August 2013. Accessed online 24 Aug 2014 at:, J. (2011) Visions of False Creek: Urban Development and Industrial Decline inVancouver, 1960-1980. MA Thesis. Department of History. University of Victoria.Moller, S. (2009) “Changing Patterns of Income Inequality in U.S. Counties, 1970-2000”American Jouranl of Sociology 114 (4): 1037-1101.119Neighbourhood Change Research Partnership (NCRP) (2014) Rental Housing DisadvantageIndex. University of Toronto.Oreopoulos, P. (2008) “Neighbourhood Effects in Canada: A Critique” Canadian Public Policy34 (2): 237-258.Oreopoulos, P. (2009) “Why Do Skilled Immigrants Struggle in the Labour Market? A FieldExperiment with Six Thousand Resumes” Metropolis Working Paper 9 (3).Osberg L (2008) A quarter century of economic inequality in Canada, 1981–2006. CanadianCentre for Policy Alternatives. Accessed online 18 July 2014 at:, A. (2010) “Shrinking middle class makes Toronto a city of socioeconomic extremes”The Globe And Mail. Toronto. Web. 25 Nov 2012.Peck, J. (2012) “Austerity urbanism” City 16: 626-655.Peck, J. and A. Tickell (2002)"Neoliberalizing space" Antipode 34 (3): 380-404.Pendakur, K. and R. Pendakur (2011) “Colour By Numbers: Minority Earnings in Canada 1996-2006” Metropolis Working Paper Series 11 (5).Piketty, T. and E. Saez (2006) “The Evolution of Top Incomes: A Historical and InternationalPerspective” American Economic Review 96 (2): 200-205.Reardon, S. and K. Bischoff (2011) “Income Inequality and Income Segregation” AmericanJournal of Sociology 116 (4): 1092-1153.Sassen, S. (1988) The Mobility of Labour and Capital: A Study in International Investment andLabor Flow. Cambridge: Cambridge University Press.Sassen, S. (2001) The Global City, Second Edition. Princeton: Princeton University Press.Sassen, S. (2006) Cities in a World Economy, 3rd edn. Thousand Oaks, CA: Pine Forge Press.Sassen, S. (2008) “Two Stops in Today’s New Global Geographies: Shaping Novel LaborSupplies and Employment Regimes” American Behavioral Scientist 52 (3): 457-496.Sinoski, K. (2013) “The new Surrey: City centre reaches for the sky” The Vancouver Sun. 2 Feb2013.Smith, H. and D. Ley (2008) “Even in Canada? The Multiscalar Construction and Experience ofConcentrated Immigrant Poverty in Gateway Cities” Annals of the Association ofAmerican Geographers 98 (3): 686-713.120Smith, C. (2012) “New Westminster’s urban revival” The Georgia Straight. 25 April 2102.Smith, C. (2013) “Towers rise in Burnaby, all according to plan” The Georgia Straight. 21 Feb2013.Stapleton, J., B. Murphy and Y. Xing (2012) The “Working Poor” in the Toronto Region: Whothey are, where they live, and how trends are changing. Report for the MetcalfFoundation. Toronto: February 2012.Timberlake, M. (2012) “Testing a Global City Hypothesis: An Assessment of Polarization acrossUS Cities” City & Community 11 (1): 74-93.TransLink (2007) Expo Line Station Review: Needs and Priority Assessment. May 2007.Accessed online 9 Sept. 2014 at: der Waal, J. and J. Burgers (2009) “Unravelling the global city Debate on Social Inequality:A Firm-level Analysis of Wage Inequality in Amsterdam and Rotterdam” Urban Studies46: 2715-2729.Van der Waal, J. (2010) Unravelling the Global City Debate: Economic inequality andethnocentrism in contemporary Dutch cities. PhD Thesis. Erasmus University Rotterdam.Volscho, T. and A. Fullerton (2005) “Metropolitan Earnings Inequity: Union and Government-Sector Employment Effects” Social Science Quarterly 86 1324-1337.Walker, L., L. Garnett, R. Kan and P. Klitz (2009) Integrating Frequent Transit Service &Corridor-based Transit Supportive Environments in the Metro Vancouver Region. Paperprepared for presentation at the 2009 Annual Conference of the TransportationAssociation of Canada. July 2009.Walks, A. (2014) “From Financialization to Sociospatial Polarization of the City? Evidence fromCanada” Economic Geography 90 (1): 33-66.Walks, R. A. and L. Bourne (2006) “Ghettos in Canada’s cities? Racial segregation, ethnicenclaves and poverty concentration in Canadian urban areas” The Canadian Geographer50(3): 273-297.Watson, T. (2009) “Inequality and the measurement of residential segregation by income inAmerican neighbourhoods” Review of Income and Wealth 55 (3): 820-844.Watt, P. (2008) “The only class in town? Gentrification and the middle-class colonization of thecity and the urban imagination” International Journal of Urban and Regional Research32: 206–211.121Wright, E.O. and R. Dwyer (2003) “The patterns of job expansions in the USA: A comparison ofthe 1960s and 1990s” Socio-Economic Review 1: 289–325.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items