打印

作者

来源

URBAN GEOGRAPHY,Vol.42,Issue2,P.181-196

语言

英文

关键字

Buy-to-let,rent gap,ethnicity,gentrification,Private Rented Sector

作者单位

a Luxembourg Institute of Socio-Economic Research (LISER), Luxembourg;b Department of Geography and Environment, London School of Economics and Political Science, London, UK

摘要

This article analyses the interrelation of ethnicity, class and tenure in the gentrification trajectories that have taken place in England in the most recent intercensal period (2001–2011). It argues that the return of the Private Rented Sector has made possible the extension of social change to areas not favored by White British (majority ethnic) middle-class owner-occupiers. This has seen the inflow of White British private renters into White British working-class areas and the arrival of private renting Not White British middle classes – primarily migrants – in working-class areas with a significant proportion of Not White British individuals. There is thus an ethnic dimension to the geographical spread of buy-to-let gentrification and the movement toward property wealth re-concentration it feeds. The middle-class individuals entering gentrifying areas as private renters are however not classical gentrifiers as the closing of rent gaps starts with the supply of private rented units by property investors.KEYWORDS: Buy-to-letrent gapethnicitygentrificationPrivate Rented SectorIntroductionThis article centers on the ethnic dimension of private rental gentrification in England. As prime inner-city locations for gentrification are increasingly exhausted, gentrification processes are starting to turn to areas with significant concentrations of disadvantage, public housing stock and/or large ethnic minority populations, characteristics posited to act as barriers to gentrification (Ley & Dobson, 2008; Walks & August, 2008).The recent return of the private rental sector in several countries such as the United States (Teresa, 2016), the Netherlands (Hochstenbach & Ronald, 2018), and the United Kingdom (Ronald & Kadi, 2018) is associated with rising property wealth inequality (Arundel, 2017). Alongside the rise of digital platforms facilitating short-term rentals, it has also diversified the types of gentrification processes occurring in cities. In Lisbon, investors and individual landlords draw on the possibilities offered by short-term rentals to replace residents by tourists, leading to the displacement of working-class communities (Cocola-Gant & Gago, 2019; see also Wachsmuth & Weisler, 2018). In Brussels, Van Criekingen (2009) writes on the role of the private rental sector in facilitating the gentrification of inner-city areas. In the case of the United Kingdom, Paccoud (2017) showed that the deregulation of the Private Rented Sector since 1988 enabled the spread of buy-to-let gentrification in more central, disadvantaged and urban areas of England than gentrification linked to increases in ownership. In London, Paccoud and Mace (2018) found that this pushed buy-to-let gentrification into suburban locations that had remained immune to previous rounds of gentrification.These changes challenge a fundamental assumption that still underpins the gentrification literature; namely that it is the middle-class individuals coming to live in disadvantaged neighborhoods who are the gentrifiers. The fact that gentrification can occur through the arrival of renters complicates this picture by introducing a third actor, the individual or collective investor who owns, and who usually has created, the dwelling that is for rent. To understand these new gentrification processes, it is, therefore, necessary to shift perspectives. Rather than focusing exclusively on the populations entering and leaving any given area – and identifying winners and losers – rental gentrification needs to be understood as an expansion of the possibilities offered by gentrification to extract value from the built environment.In considering this extraction of value we draw on the concepts of both “rent gap” and “value gap”. While Hamnett and Randolph (1984) developed the concept of a “value gap” to explain an earlier shift of tenure from the Private Rented Sector to owner occupation, we draw on its underexploited value as an explanation for the reverse movement as fundamentally it describes the logic of investors driving tenure change to realize the highest returns. However, as Paccoud (2017) argues, the risks to rewards for small-scale buy-to-let landlords are not sufficiently clear to explain the mass take up of buy-to-let using the “value gap” as an explanation. The appeal of buy-to-let is that it speeds up the process by which investors can close the gap in value between current and future use – what has been called the rent gap (Smith, 1979), including through a high churn of tenants once a property is on the rental market. Drawing on Clark’s (1992) attempt to integrate value and rent gaps, buy-to-let gentrification can thus be thought of as a “response to the generalised creation of value gaps which provides local opportunities for investors to at least partially close the rent gap by introducing middle class renters into disadvantaged but central areas” (Paccoud, 2017).To better elicit the mechanisms which structure these local opportunities for rent gap closure, this paper focuses on the ways in which buy-to-let gentrification relates to ethnicity, both in terms of the spatial distribution of ethnic groups and in terms of the types of populations involved in social change. While there are differences in how racial (and class) politics play out in the United States and United Kingdom, evidence from both nations shows that areas with a significant minority population are likely to be less attractive to White majority gentrifiers (Bader & Krysan, 2015; Charles, 2003; Timberlake & Johns-Wolfe, 2017). There are significant differences between gentrifiers (Butler & With Robson, 2003; Jackson & Benson, 2014; McKinnish, Walsh, & White, 2010), with some more likely to accept or positively value ethnically diverse neighborhoods. However, the overall tendency of White gentrifiers to avoid minority ethnic areas means that despite large rent gaps in some inner-city locations (Lees, 2014; Watt, 2009), the pressure for value extraction from these areas has not been met by arrivals of ethnic majority (White British) owner occupiers. We show that the return of the Private Rented Sector is providing a way to circumvent the difficulties inherent in the gentrification of these areas.We identified three configurations of social, tenure and ethnic change that co-existed in England between 2001 and 2011. The first is the replacement of working-class White British owners by middle-class White British owners in White British working-class areas. This type of ownership gentrification is the benchmark: these are the areas in which middle-class individuals of the majority ethnicity freely move in as owner-occupiers. As such, it most closely mirrors Glass’ (Glass, 1988) conception of the term.The second type occurs through the arrival of White British middle-class private renters in areas in which the working class was also near exclusively White British. Here, it is the tenure shift that propels capital accumulation in areas not exploitable through ownership gentrification.Finally, the third type involves the arrival of Not White British middle-class private renters in areas in which a large proportion of the working class was Not White British in 2001. Here, the shift to the Private Rented Sector is not enough to entice middle-class individuals of the majority ethnicity to move in. It is skilled migrants looking for a central city location that provide a means to close rent gaps.The move from the first to the third of these configurations is associated with the exploitation of progressively larger rent gaps. This gradation in the size of rent gaps, empirically verified at the end of the analysis that follows, depends on the opportunity afforded by buy-to-let gentrification to ratchet up potential rents in places in which both actual and potential rents seemed destined to remain low because they were considered “off-limits” for ownership. This opportunity to ratchet up potential rents is felt most strongly in places that have been most off-limits to traditional gentrification: areas with significant ethnic minority populations – having resisted gentrification for longer, the ability of buy-to-let gentrification to push potential rents up is increased in these inner-city areas compared to areas more accessible to White British middle-class individuals.Our results confirm what other studies have shown: the characteristics of any given low-income population which occupy a neighborhood – and how it is perceived socially – plays a central role in the type of gentrification that takes place there, as well as determines who the in-movers are likely to be. Two mechanisms are at work here: first, the middle-class White British population which generally does not want to purchase in working-class areas (regardless of the ethnic composition) might be open to renting there and; second, while the middle-class White British population may not want to rent in ethnic minority, working-class areas, middle-class migrants might. These results point to the fact that buy-to-let gentrification diminishes the extent to which either the White British population’s preference for ethnically matched areas, or the co-concentration of ethnic minority and deprived areas, acts as a barrier to inward flows of middle-class individuals. This is leading to a spread of gentrification processes to new areas, and thus to a large-scale social transformation of “ordinary” areas with a co-concentration of a significant ethnic minority and deprived population, in opposition to more local and specific instances of White British owner-occupiers gentrifying minority ethnic areas.Buy-to-let gentrification in the United KingdomA significant facilitator of private rental gentrification in the United Kingdom was the 1988 Housing Act which provided a more permissive environment for landlords, including easing repossession of a property, helping to precipitate the return of the Private Rented Sector. This was further aided by the introduction of buy-to-let mortgage products in 1996 which assisted people in purchasing a property to rent out. In the early twentieth-century three-quarters of households were in the Private Rented Sector but by 1991 the sector represented just 9% of households. By 2001 and the proportion was 12% rising to 18% in 2011 (ONS, 2013), an increase of 1.7 million households.This article is a development of the study of two types of gentrification described in Paccoud (2017) and Paccoud and Mace (2018). Therefore, in the rest of this section, we provide an outline of the general method before, in the next section, turning to the ethnic dimension that is the subject of this article.The method draws on the analysis of the 161,296 Output Areas in England whose borders remained unchanged between 2001 and 2011. Output Areas or OAs, have a population of around 300 in England, making them the smallest geography for which census data are available1. An OA is considered to have experienced gentrification between 2001 and 2011 if it displays a social upscaling trajectory. This is in line with much gentrification research (Hochstenbach & van Gent, 2015; Ley & Dobson, 2008; Timberlake & Johns-Wolfe, 2017; Walks & Maaranen, 2008), but what is new here is the crossing of social with tenure change, which reveals two different forms of gentrification: (1) Ownership gentrification is defined as social upscaling accompanied by a maintenance of an owner-occupier majority or the transfer of dwellings to ownership from other tenures; (2) Buy-to-let gentrification occurs when a social upscaling trajectory is accompanied either by a shift of dwellings from owner-occupation or social renting to private renting or by an intensification of private renting.Social upscaling is represented through the Office for National Statistics’ “Socio-economic Classification” (hereafter NS-SeC). The dataset allocates a key household member (Household Reference Person) aged 16 to 64 to one of eight broad occupational categories: [1] higher managerial, administrative and professional occupations; [2] lower managerial, administrative and professional occupations; [3] intermediate occupations; [4] small employers and own account workers; [5] lower supervisory and technical occupations; [6] semi-routine occupations; [7] routine occupations; and [8] never worked and long-term unemployed. The 11,859 OAs in which social upscaling was identified between 2001 and 2011 had arrivals of members of the first two NS-SeC categories and departures of members of categories 5, 6 and 7 that were larger than a threshold set at 7% of the total number of Household Reference Persons aged 16–64 in the Output Area in 2001, with groups 3 and 4 co-varying with either arrivals or departures depending on particular conditions (Paccoud, 2017; Paccoud & Mace, 2018, p. 1316). The same threshold of 7% of the 2001 OA usual resident population was used to identify the tenure trajectories that have occurred in these upscaling OAs, with tenure types aggregated into three main categories: owners, social renters and private renters.Given our use of the rent gap, we draw on evidence that occupational changes are related to ground rent changes. While the methodology relies on occupational change in small areas, there is a lack of very local information on economic phenomena linked to land and housing markets. Given this, we draw on official statistics that provide the median price paid for residential properties by Lower Super Output Area (LSOA), the next smallest statistical area after the Output Area (OA)2. For England as a whole, the 2,883 LSOAs in which at least 20% of OAs had experienced occupational upscaling (as defined above) between 2001 and 2011 experienced larger house price increases between 2001 and 2011 than the 2,643 LSOAs in which at least 20% of OAs had seen occupational downscaling (the replacement of NS-SeC groups 1 and 2 by NS-SeC groups 5, 6 and 7). This is true in each of the nine regions of England, though the differences are modest (between +0.2% in Yorkshire and the Humber and +5.4% in the South West for upscaling compared to downscaling LSOAs). This is likely a result of the drowning out of OA level changes at the LSOA level. Indeed, larger differences are found for stricter LSOA criteria: between +0.6% in the East of England and +9% South West of England for the 1,146 LSOAs with at least 40% of upscaling OAs compared to the 959 LSOAs with at least 40% of downscaling OAs (beyond this level, the number of LSOAs becomes too small to allow a regional analysis). It is also possible that this is linked to the impact of the 2008 recession on house prices given that these differences are larger for the longer 1997 to 2017 period (between +1.3% for the South East and +9.6% for Yorkshire and the Humber for the “at least 20% of OAs per LSOA specification”). In the absence of more precise information, this crude analysis does provide some indication that occupational and house price changes go in the same general direction.The method of identifying gentrification described above relies on the association of different dimensions of Output Area level changes (occupation, tenure and ethnic change), rather than on cross-tabulations. The criteria used to identify these changes have been made strict enough to guarantee that the identified OAs feature changes in population characteristics that are too large to have occurred independently in three different dimensions. The method may potentially capture some generalized in situ upward shifts in the NS-SeC distribution – in other words upscaling without residential movement. But, a significant number of Household Reference Persons would have had to shift from NS-SeC categories 5, 6 and 7 to categories 1 and 2 within a 10-year period. There is some numerical evidence that this has not occurred on a large scale in England between 2001 and 2011. Among the 11,859 OAs defined as upscaling, only 136 OAs (or 1.2% of all upscaling OAs) had not experienced any significant change in either their age structure, ethnic structure or country of birth profile.The analysis that follows will zoom in on 4,387 Output Areas out of the 11,859 with social upscaling, as in these OAs this upscaling has been accompanied by particular tenure changes. Ownership gentrification accounts for 1,850 of these OAs and can occur through three main tenure trajectories: minimal tenure change in OAs in which home-owners were the majority in 2001 (957 OAs), a sole increase of owners (677 OAs) and the replacement of social tenants by individuals owning (216 OAs). Buy-to-let gentrification – which accounts for 2,537 OAs – is also represented by three major tenure trajectories: the replacement of owners by private renters (1,451 OAs), the sole increase in private renters (801 OAs) and the replacement of social by private renters (285 OAs). As discussed in more detail in Paccoud (2017), around 45% of social upscaling OAs featured tenure changes that were too small or too mixed to be clearly categorized.The ethnic dimension of buy-to-let and ownership gentrificationDifferent from the earlier cited studies focused on the White/Black dimension of gentrification our first level of distinction is between the majority ethnic group (White British), and all other ethnic groups, i.e. White minority as well as Black minority ethnic groups (Not White British). We also distinguish between first-generation migrants (born abroad) and the rest as this yields significant results.3A first glance at the net population changes that have accompanied buy-to-let and ownership gentrification between 2001 and 2011 reveals a clear difference between these two types of gentrification when it comes to ethnicity and country of birth. In the 2,537 Output Areas which experienced buy-to-let gentrification between 2001 and 2011, there was a net departure of 36,205 White British persons and the net arrival of 56,012 persons of other ethnicities, among which one-third was born in the United Kingdom and the other two-thirds born outside of the United Kingdom. The biggest net increases were among the White Other population born overseas (mostly born in European countries outside of the EU154), those of Asian ethnicity born outside the United Kingdom (who mostly self-identify as Other Asian, Indian or Pakistani), as well as those of Asian or mixed ethnicity who were born in the United Kingdom. When related to the number of OAs involved in this type of gentrification, it becomes even clearer that the main population movement in these areas was the replacement of White British by individuals of other ethnicities: per OA on average, there was a decrease of −14.3 White British alongside a net increase of +22.1 individuals from other ethnicities.In contrast, the 1,850 Output Areas with ownership-led gentrification between 2001 and 2011 saw a net increase of 7,732 White British persons, alongside a net increase of 17,430 persons of other ethnicities, with a slight majority of these individuals born in the United Kingdom. The White British are those with the largest net increase in population over this period, followed by Asians born overseas, Asians born in the United Kingdom and individuals of mixed ethnicity born in the United Kingdom. When related to the number of OAs concerned, the situation is more difficult to read than for buy-to-let gentrification. Indeed, the average net increase in the White British (the ethnicity with the single largest net increase) was of only 4.2 individuals per OA, and the net average increase of all other ethnicities added up to only 9.4 individuals per OA. The ethnic profile of these OAs has thus changed much less and it is not immediately clear which group has left the area through the process of social and tenure change. Given the strict criteria used to identify OAs in which gentrification took place, a large population change necessarily occurred in these areas through the replacement of individuals of low NS-SeC occupations by individuals with high NS-SeC occupations. However, no ethnic group has seen a strong decline in those OAs between the two Censuses, apart from a slight net decrease in the Irish born in the United Kingdom. It is thus likely that there has been a replacement of individuals within each ethnic group.It thus seems as though there are two different ethnicity-related logics in areas having undergone gentrification linked to the Private Rented Sector and owner-occupation. In order to get a better grasp of these complex processes, we first provide two perspectives on the ethnic dimension of buy-to-let and ownership gentrification: the ethnic make-up of the working-class areas targeted and the local trajectories of ethnic change through which they occurred. We then cross these two perspectives to identify three configurations of social, tenure and ethnic change. In a final section, we present evidence that the configurations that draw on shifts to the Private Rented Sector and that occur through the arrival of middle-class Not White British individuals have had the largest housing market impacts.The context prior to gentrification: ward ethnic and occupational profiles in 2001In this first step, we distinguish between two types of wards, areas larger than Output Areas, in 2001, by combining occupation (as a proxy for class) and ethnicity data from the 2001 Census5: NWBS+WC (Not White British Significant + Working Class) wards; in which more than 2.5% of the ward’s population was made up of Not White British individuals in NS-SeC groups 5, 6 and 7 (the average figure for all wards is 1.3%) and over 15% of those in NS-SeC groups 5, 6 and 7 in these wards were Not White British (the average figure for all wards is 7.6%). The 847 out of 7,952 wards (10.6%) which fit these criteria are predominantly in Central, North and West London and in the central parts of other English cities (including Birmingham and Leeds). In 2001, these wards had 17.4% of their population in groups 5, 6 and 7;WBD+WC (White British Dominant + Working Class) wards; in which more than 15% of the ward’s population was made up of White British individuals in NS-SeC groups 5, 6 and 7 (the average figure for all wards is 18.6%) and over 95% of those in NS-SeC groups 5, 6 and 7 in these wards were White British (the average figure for all wards is 92.4%). The 4,203 wards in this category represented 53% of all wards in the United Kingdom in 2001 and can be found across all of England, with the exception of the greater South East. In 2001, these wards had 22.8% of their population in groups 5, 6 and 7.In this section, we only focus on the Output Areas with buy-to-let or ownership gentrification that are located in either of these two types of wards (1,690 out of 2,537 OAs in the case of buy-to-let gentrification and 1,048 out of 1,850 OAs for ownership gentrification). Table 1 reveals clear differences in the ethnic make-up of the working-class areas in which these two types of gentrification occurred.The role of ethnic change in the closing of rent gaps through buy-to-let gentrificationAll authorsAntoine Paccoud , Pauline Niesseron & Alan Macehttps://doi.org/10.1080/02723638.2019.1709755Published online:03 January 2020Table 1. Average net change of White British and Not White British in buy-to-let and ownership gentrification (by ward type)CSVDisplay Table The wards in which there was a large number of working-class individuals of all ethnicities other than White British in 2001 (NWBS+WC wards) had a replacement of White British by Not White British in Output Areas with gentrification both to the Private Rented Sector and to ownership. In the 618 OAs with buy-to-let gentrification located in the first type of ward, the largest net decrease was in the White British population (−28.3 individuals in each), and the largest increases were in the White Other population born overseas (with an average increase of 15.2 individuals per OA, not shown in Table 1) and Asians born overseas (+13.6 individuals per OA, not in Table 1). This picture is similar in its broad strokes to that given for buy-to-let gentrification OAs in England more generally but the degree of change is greater here, with double the departure of White British (−28.3) and double the increase in individuals of all other ethnicities (+46) per OA.Still in the NWBS+WC wards, the 112 Output Areas in which ownership gentrification took place saw the same large decrease in the White British population (−27.8 individuals per OA), but the largest increases here are found among those with Asian ethnicity born in the United Kingdom (+15.2 per Output Area, not shown in Table 1) as well as born overseas (+12.7 per OA, not shown in Table 1); the majority of which seem to be of Pakistani and Indian ethnicity. Thus, while the White British are the group with the largest net increase into OAs with ownership gentrification as a whole, they are the largest leavers by far in the small subset of this type of gentrification in NWBS+WC wards.Moving to gentrification which occurred in wards in which the working class was near exclusively White British provides the other side of the picture (WBD+WC wards). In the case of buy-to-let gentrification, the intensity of change in the 1,072 Output Areas in this type of ward is halved as compared to all 2,537 Output Areas with buy-to-let gentrification in England (−7.6 White British and +8.4 individuals of other ethnicities per OA) and no particular Not White British group experienced significant population movements. This points to a replacement of individuals of low NS-SeC by those of high NS-SeC occupations within given ethnic groups.In the 936 Output Areas in which ownership gentrification took place in wards where the working class was near exclusively White British, the intensity of change remains as low as for all 1,850 ownership gentrification OAs, with the only notable change being the fact that there are larger net increases in the White British population (+7.6 per OA) than in the population of other ethnicities (+4.2 per OA), a reversal of the situation in all ownership gentrification OAs. Again, this indicates that social change within given ethnic groups is likely to have occurred here.The focus on the ethnic make-up of the working-class areas that gentrified between 2001 and 2011 seems to show that the gentrification landscape is split along ethnic lines: while Not White British individuals are heavily involved in the gentrification of NWBS+WC wards, they are much less present in the gentrification of wards where the working class was near exclusively White British. To get abetter sense of the processes at work in gentrifying areas, and especially in those in which there was a low intensity of change at the aggregate level, the next section shifts the focus to the categorization of ethnic change at the Output Area level.The population shifts accompanying gentrification: categories of local ethnic changeTo gain a different perspective on the ethnic dimension of buy-to-let and ownership gentrification, we focus on a categorization of ethnic change at the Output Area level. Using the same method as for social and tenure change (which considers that a change has occurred if arrivals and departures represent over 7% of the 2001 OA resident population), we found five population trajectories that covered 3,105 out of the 4,387 OAs in which either buy-to-let or ownership gentrification occurred between 2001 and 20116. The rest of this section will thus focus on these 3,105 OAs and the ways in which the trajectories of local ethnic change are distributed between buy-to-let and ownership gentrification. To better observe the main trends, these five major categories can be combined into three (A–C as follows):A. Movements within the White British population. These are the OAs with a sole increase of White British individuals, a sole decrease of White British individuals or areas with a White British majority in 2001 in which there has been minimal population change between 2001 and 2011. In these three types of OAs (i, ii and iii), large social and tenure changes are accommodated within the White British population. White British up (572 OAs): the sole increase of White British individuals – here the increase in high NS-SeC WB masks the departure of low NS-SeC White British individuals.White British down (475 OAs): the sole decrease of White British individuals – here the departure of low NS-SeC White British individuals overshadows the increase in high NS-SeC White British individuals.Minimal population change in White British majority areas (851 OAs) – here there was a replacement of low NS-SeC White British by high NS-SeC White British.B. Movements within all other ethnicities (Not White British). The Output Areas in which population movements are accommodated within the population of Not White British ethnicity – the only significant trajectory here is the sole increase of Not White British individuals.Not White British up (471 OAs): the sole increase of Not White British individuals – here the increase in high NS-SeC Not White British masks the departure of low NS-SeC Not White British, and;C. Not White British replace White British. OAs where there is also a single significant population change, the replacement of the White British by other ethnicities.White British down, Not White British up (736 OAs): the replacement of low NS-SeC White British by high NS-SeC Not White British.The figures in Table 2 – which maps the local ethnic change trajectories on the two types of gentrification that occurred between 2001 and 2011 – confirm the differences in terms of population change associated with buy-to-let and ownership gentrification. While the Not White British have played an important role in buy-to-let gentrification (with OAs in categories B and C representing 52.4% of such gentrification, compared to 47.6% for category A), they are heavily under-represented in ownership gentrification, with categories B and C making up only 19.7% of such gentrification (as compared to 80.3% for category A).The role of ethnic change in the closing of rent gaps through buy-to-let gentrificationAll authorsAntoine Paccoud , Pauline Niesseron & Alan Macehttps://doi.org/10.1080/02723638.2019.1709755Published online:03 January 2020Table 2. The distribution of the three main ethnic trajectories in ownership and buy-to-let gentrification Output AreasCSVDisplay Table Identifying three gentrification configurations: ward profiles and ethnic trajectoriesThe analysis has so far looked at the ethnicity-related differences between buy-to-let and ownership gentrification from two perspectives: first, through the ethnic make-up of the working-class areas in which each type of gentrification occurred; and second, through the local trajectories of ethnic change that have accompanied them.This section focuses on the crossing of these two perspectives, with the aim of drawing out the three broad configurations of social, tenure and ethnic change that took place in England between 2001 and 2011. This means reducing the number of gentrifying Output Areas once again, since only 1,953 OAs fulfill all the criteria: they are located in either NWBS+WC or WBD+WC wards and have gentrified through one of the three broad trajectories of ethnic change discussed in the previous section. Of these remaining 1,953 OAs, 739 experienced ownership gentrification and 1,214 buy-to-let gentrification. These will be used to distinguish between three gentrification configurations.The first configuration (type 1) is the replacement of working-class White British owners by middle-class White British owners in wards in which the working class was near exclusively White British in 2001. It can be clearly seen in Table 3, which focuses on the distribution of the 739 OAs (out of the 1,953) with ownership gentrification. The vast majority of this gentrification occurred in wards in which the working class was near exclusively White British (650 OAs as compared to 89 OAs for the wards with large Not White British working-class populations). Moreover, most of the social change in these wards occurred within the White British population (608 OAs as compared to 42 OAs for the other ethnic trajectories).The role of ethnic change in the closing of rent gaps through buy-to-let gentrificationAll authorsAntoine Paccoud , Pauline Niesseron & Alan Macehttps://doi.org/10.1080/02723638.2019.1709755Published online:03 January 2020Table 3. Crossing of ward type and ethnic trajectory for outcomes leading to ownership gentrificationCSVDisplay Table The two other configurations of social, tenure and ethnic change are found within the 1,214 Output Areas (out of the 1,953) with buy-to-let gentrification and are shown in Table 4. The second configuration (type 2) took place through transfers of dwellings to the Private Rented Sector in areas in which the working class was also near exclusively White British. It was propelled by the replacement of working-class White British by middle-class White British individuals. It is visible in the fact that the vast majority of the buy-to-let gentrification that occurred in the wards in which the working class was near exclusively White British was driven by inflows of middle-class White British renters (565 Output Areas vs 122 in which the incoming private renters are Not White British). Compared to type 1, in which White British buy into working-class areas, it is the tenure shift to the Private Rented Sector (either from social renting or from ownership) which enables the inflow of middle-class White British individuals.The role of ethnic change in the closing of rent gaps through buy-to-let gentrificationAll authorsAntoine Paccoud , Pauline Niesseron & Alan Macehttps://doi.org/10.1080/02723638.2019.1709755Published online:03 January 2020Table 4. Crossing of ward type and ethnic trajectory for outcomes leading to private rental gentrificationCSVDisplay Table The final, third configuration (type 3) occurred through the arrival of Not White British middle-class renters in places in which a large proportion of the working class was Not White British. This can be seen in Table 4 by the fact that of 527 Output Areas located in these areas in which buy-to-let gentrification took place, 470 can be linked to inflows of middle-class renters of ethnicities other than White British and only 57 to inflows of White British middle-class renters. Here, the shift to the Private Rented Sector is not enough to entice White British middle-class individuals to move in and gentrification can only spread through the inflow of middle classes of other ethnicities. The largest net population increases are among those of White Other and Asian ethnicity born overseas. It is thus middle-class migrants looking for a central city location while they make their first steps in the United Kingdom that provide a means to close rent gaps in the wards with large Not White British working-class populations.Assessing the housing market impacts of the three configurationsIt was hypothesized at the top of article – drawing on evidence from the gentrification literature – that rent gaps would be larger in places that had resisted gentrification for longer. These are places with significant concentrations of disadvantage, public housing stock and/or large ethnic minority populations. Buy-to-let gentrification was identified as a mechanism through which social upscaling could spread to these areas, leading to the closure of larger rent gaps than in areas targeted by (largely White British) ownership gentrification.We are now able to put this hypothesis to the test by comparing the housing market impacts of the three gentrification configurations identified in the last section. To do this, we turn once again to median residential property prices at the Lower Super Output Area (LSOA) level and compare the evolution of these prices for LSOAs containing Output Areas with each of the three gentrification configurations. While there is some evidence that there are indeed differences between these three configurations, the comparison is made difficult by their different regional compositions: types 1 and 2 are fairly evenly distributed across all regions except London, while two-thirds of type 3 is found in London (and none of it is found in the North East). It is thus only possible to compare house price changes in seven out of the nine regions for all three types. Two results do however stand out.First, for two specifications7 and for two time periods (2001–2011 and 1997–2017), it is generally the case that LSOAs containing OAs of types 2 and 3 saw higher house price increases than LSOAs with type 1 OAs. The only exceptions are for two southern regions (the South East and the East of England) for type 3 in 2001–2011 and two northern regions for type 2 in 1997–2017. Second, across the same specifications and time periods, LSOAs containing type 3 OAs tend to have had larger house price changes than those with type 2 OAs in the seven regions where this comparison is possible. The only exception – similarly to above – is for the South East and the East England between 2001 and 2011.A more detailed analysis of house price change under the three configurations of social, tenure and ethnic change is hindered by the lack of house price information at the OA level and the relatively small number of observations for type 3 when split regionally. There is however some indication that buy-to-let gentrification (which underlies types 2 and 3) is linked to higher house price increases, a possible reflection of its ability to ratchet up potential rents by bypassing some of the barriers faced by the “historical” gentrification configuration. That is, the arrival of White British middle-class owner-occupiers in inner-city working-class areas home to social renters and/or Not White British working-class populations.DiscussionIn this research, gentrification has been identified by looking at the replacement of low socioeconomic status residents by high socioeconomic status residents, facilitated by a tenure change to either the Private Rented Sector or to owner-occupation, between the 2001 and 2011 censuses. The aim of the analysis has been to take stock of the implications of a new type of gentrification linked to the Private Rented Sector for the types of areas impacted by the process. The results provide a systematic look at the way in which gentrification intersects with tenure, ethnicity and migration. Our approach using Census data to record patterns of movement along a simple division of White British and all other ethnicities focuses on outcomes. While we present the most likely pathways through which gentrification intersected with ethnicity, we acknowledge that we cannot gain insight into more locally grounded dynamics. We thus cannot discuss the role of racism, both exercised and experienced, in informing the data we employ here. This is an important limitation given the evidence that racism impacts on decisions of where to move. For example, in the case of British Asians moving to suburban areas of Leeds and Bradford, Phillips, Davis, and Ratcliffe (2007) provide an account of how the choice of a new neighborhood is guided by readings of the likelihood of being subject to racism. Investigating gentrification in the United States, Wyly and Hammel (2004) focus on discrimination toward people of color seeking to buy into gentrifying areas.Acknowledging this limitation, we have sought to build on our previous work to point to the centrality of the ethnic dimension in the spatial expansion of gentrification that occurs through the Private Rented Sector. In doing so, we seek to respond to the critique that work on gentrification has in the past paid insufficient attention to ethnicity. When it has, contra intersectionality, the gentrification literature has oftentimes treated ethnicity as nothing more than a correlate of class, leading to a call to go beyond, “[the] implicit race and class oppositions organizing the gentrification literature: middle-class gentrifiers/incomers (White) versus working-class residents/displaced (Black)” (Lees, 2000, p. 400). As Moore (2009) later showed, “race and ethnicity shape the processes of gentrification in more complex ways … ” (118). The data presented here indeed provide a more complex picture than the one usually painted of the interconnection of gentrification and ethnicity. None of the three broad configurations of social, tenure and ethnic change described involve the replacement of Not White British by White British middle-class individuals. In fact, there is only a limited number of Output Areas that followed the usual White replaces Black narrative. These are the ones in which there was the replacement of some working-class individuals from a given ethnic minority group by middle-class individuals of other another ethnicity, especially through the replacement of social and private renters by private renters in minority ethnic working-class wards. In this subset of type 3 Output Areas (those with the population trajectory Not White British up), there are net decreases of individuals of Black Caribbean and White Irish ethnicity, with other groups showing relatively small intensities of change. The large inflows of those of the ethnicity White Other in these areas is what looks most like the traditional narrative, except that the middle-class minority renters are only the vehicles through which gentrification is pushed forward, not the engines of the process. These incoming middle-class minority ethnic private renters are not gentrifiers in the classical sense of the term since they had no role in the production of the new private rental dwellings and in their price level. The disconnect between the in-movers and the process of gentrification is thus pushed one step further in buy-to-let gentrification as compared to “classical gentrification”, in the context of which Rose (1984) coined the term “marginal gentrifier” to indicate the variety of class positions that in-movers might hold. This disconnect pushes for an interpretation of gentrification more in line with Smith’s (1987) response to the challenge of the “marginal gentrifier”: buy-to-let gentrification entails a clear polarization not only between the investor and the displaced but also between the investor and the tenants who participate in the process.These results give credence to the fact that the return of the Private Rented Sector has ratcheted up potential rents in places that had posed barriers to ownership gentrification. The closure of these rent gaps occurred through the inflow of ethnically differentiated middle-class renters depending on the ethnic profile of the working-class population living in the area: White British renters in working-class areas that are near exclusively White British; middle-class migrants in working-class areas that have large proportions of Not White British individuals. This contrasts with the more stable geography of ownership gentrification, driven by movements within the White British population in wards in which the working class was near exclusively White British. The concept of “territorial stigmatisation” mobilised by Wacquant, Slater, and Pereira (2014) could help explain the relative absence of White ownership gentrification in areas with a significant ethnic minority population: the negative and stereotypical media representation of ethnic minority areas might result in those areas being less attractive to the ethnic majority group. While the spatial extension of gentrification seems to be feeding into local house price growth, it remains to be seen whether this will lead to a long-term upgrade of the built environment or if it represents an opportunistic mechanism to profit from potential rent increases. The future impacts of the phenomenon we describe is the subject of work in the United States by Teresa (2016) who focuses on the actions of buy-to-let landlords seeking to realize the hope value of their investment.The data reinforce the literature that highlights that gentrification in minority working-class areas tends to occur, in the first instance, through the arrival of co-ethnics or migrants rather than through the inflow of middle-class White British individuals. This is reflective of the experience of gentrification in some of the largest US cities (Anderson & Sternberg, 2013; Bostic, Martin, & Richard, 2003; Boyd, 2005; Pattillo, 2003) or Canada’s largest cities (Walks & Maaranen, 2008); immigrants tend to establish strong ethnic networks commonly achieved through a place-based social fabric (Betancur, 2011) and White residents prefer White majority ethnic areas (Timberlake & Johns-Wolfe, 2017). In the English case described here, the extension of gentrification to Not White British working-class areas leaves levels of ethnic segregation untouched at the same time as it intensifies displacement pressures.To conclude, buy-to-let gentrification is shown to have opened wide avenues for gentrification in Not White British (significant) working-class places, areas that had been virtually untouched by ownership gentrification. This type of gentrification occurs through the inflow of middle-class migrants, likely attracted to these central city areas by their proximity to employment centers. This spatial extension of gentrification processes and the associated closing of new and large rent gaps is facilitated by the high population turnover in English cities. In a sense, the United Kingdom's urban super-diversity is hostage to the process of property wealth concentration signaled by the return of the Private Rented Sector. Middle-class private renters (both White British and Not White British) are only cannon fodder for a broader process of buy-to-let gentrification they have no control over. In buy-to-let gentrification, while there may be superficially different gentrifiers in different parts of the country, at the end it is the same category who propel the process (owners who sell in working-class areas, buy-to-let investors, property developers), and the same category which loses out: social and private renters who are either directly displaced or who remain in gentrifying areas and face displacement pressures.Table 1. Average net change of White British and Not White British in buy-to-let and ownership gentrification (by ward type)OutputArea (OA)withAverage net change per OA between 2001 and 2011 forNWBS+WC 2001Not White British Significant + Working ClassWBD+WC 2001White British Dominant + Working ClassAverage across all OAsBuy-to-let gentrification between 2001and 2011 618 OAs1,072 OAs= 1,690 OAsWhite British (persons)−28.3−7.6−14.3Not White British (persons)+46+8.4+22.1Ownership gentrification between 2001and 2011 112 OAs936 OAs= 1,048 OAsWhite British (persons)−27.8+7.6+4.2Not White British (persons)+42.4+4.2+9.4Table 2. The distribution of the three main ethnic trajectories in ownership and buy-to-let gentrification Output AreasBetween 2001 and 2011,OAs with →Population change linked to ↓Buy-to-let gentrificationOwnership gentrificationA. Movements within the White British populationi. White British up177395ii. White British down361114iii. Minimal population change in White British majority areas328523A. by Output Area; #/total866/18201,032/1285A. by %47.6%80.3%B. Movements within the Not White British populationNot White British up348123C. Not White British replace White BritishWhite British down, not White British up606130B. and C. by Output Area; #/total954/1820253/1285B. and C. by %52.4%19.7%Table 3. Crossing of ward type and ethnic trajectory for outcomes leading to ownership gentrification2001 →2001–2011↓NWBS+WCNot White British Significant+ Working ClassWBD+WCWhite British Dominant+ Working ClassA. White British replace White British14/89608/650‘type 1ʹB. Not White British replace Not White British20/8928/650C. Not White British replace White British55/8914/650Table 4. Crossing of ward type and ethnic trajectory for outcomes leading to private rental gentrification2001 →2001–2011↓NWBS+WCNot White British Significant+ Working ClassWBD+WCWhite British Dominant+ Working ClassA. White British replace White British57/527565/687“type 2”B. Not White British replace Not White British145/527“type 3”53/687C. Not White British replace White British325/527‘type 3ʹ69/687AcknowledgmentsWe would like to thank the three anonymous reviewers whose comments have helped improve an earlier draft; also, Ian Gordon for his comments.Disclosure statementNo potential conflict of interest was reported by the authors.Notes1. The 2011 Output Areas have an average population of 309 persons (297 in 2001), and have strict upper and lower population thresholds: they are split if they contain more than 625 persons and are merged if they contain less than 100. Between 2001 and 2011, 2.6% of OAs were merged or split. The UK Office for National Statistics reports that OAs were designed to have similar population sizes.2. An LSOA is a collection of OAs – there are 32,598 LSOAs in England and Wales, with an average of 1,500 residents. An LSOA thus contains 5 OAs on average. The dataset is the UK Office for National Statistics’ House Price Statistics for Small Areas (HPSSAs), dataset 46.3. Drawing on the univariate tables for ethnicity and country of birth in the 2001 and 2011 UK Censuses as well as the tables LC2205EW, Country of birth by ethnic group (2011 Census, persons) & CT003, Theme table on ethnic group – people (2001 Census, persons).4. The 15 member countries in the European Union prior to the accession of 10 candidate countries on 1 May 2004. The EU15 comprised the following countries: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain, Sweden, United Kingdom.5. These are the 2003 standard wards, the smallest geography for which it is possible to have Census 2001 data on the NS-SeC of ethnic groups (ST112 – Sex and NS-SeC by ethnic group). In England and Wales, these wards have an average population of 4,000.6. Of the other 1,282 Output Areas, 1,193 could not be linked to any particular trajectory of ethnic change and the remaining 89 are scattered across five minor types of ethnic change.7. Selecting only LSOAs in which (1) at least 20% of OAs or (2) more than 20% of OAs exhibit a particular type of gentrification.ReferencesAnderson, Matthew. B., & Sternberg, Carolina (2013) “Not-White” Gentrification in Chicago’s Bronzeville and Pilsen: Racial Economy and the Intraurban Contingency of Urban Redevelopment. Urban Affairs Review, 49(3), 435–467. [Crossref], [Web of Science ®], [Google Scholar]Arundel, Rowan (2017) Equity inequity: Housing wealth inequality, inter and intra-generational divergences, and the Rise of Private Landlordism. Housing, Theory and Society, 34(2),176–200. [Taylor & Francis Online], [Web of Science ®], [Google Scholar]Bader, Michael, D. M., & Krysan, Maria (2015) Community Attraction and Avoidance in Chicago. The ANNALS of the American Academy of Political and Social Science, 660(1), 261–281. [Crossref], [Web of Science ®], [Google Scholar]Betancur, John (2011). Gentrification and community fabric in Chicago. Urban Studies, 48(2), 383–406. [Crossref], [PubMed], [Web of Science ®], [Google Scholar]Bostic, Raphael., &, W. Martin, Richard, W. (2003) Black home-owners as a gentrifying force? Neighbourhood dynamics in the context of minority home-ownership. Urban Studies, 40(12), 2427–2449. [Crossref], [Web of Science ®], [Google Scholar]Boyd, Michelle (2005) The downside of racial uplift: Meaning of gentrification in an African American neighborhood. City & Society, 17(2), 265–288. [Crossref], [Google Scholar]Butler, Tim, & With Robson, Gary. (2003). London Calling: The Middle Classes and the Remaking of Inner London. Oxford: Berg Publishers. [Google Scholar]Charles, Camille Zubrinsky (2003) The Dynamics of Racial Residential Segregation. Annual Review of Sociology, 29(1), 167–207. [Crossref], [Web of Science ®], [Google Scholar]Clark, Eric. (1992) On gaps in gentrification theory. Housing Studies, 7(1): 16–26. [Taylor & Francis Online], [Google Scholar]Cocola-Gant, Agustin, & Gago, Ana (2019) “Airbnb, buy-to-let investment and tourism-driven displacement: A case study in Lisbon.” Environment and Planning A. Advance Online Publication, doi:10.1177/0308518X19869012. [Crossref], [Web of Science ®], [Google Scholar]Glass, Ruth. (1988). London: Aspects of change. In R Glass (Ed.), Clichés of Urban Doom and Other Essays (pp. 133–158). Oxford: Basil Blackwell. [Google Scholar]Hamnett, Chris, & Randolph, Wallace Alan (1984) The role of landlord disinvestment in housing market transformation: An analysis of the flat break-up market in central London. Transactions of the Institute of British Geographers, 9 (3)259–279. [Crossref], [Web of Science ®], [Google Scholar]Hochstenbach, Cody, & Ronald, Richard (2018) The revival of the private rental sector under Amsterdam’s regulated marketization regime. Centre for Urban Studies Working Paper Series No. 34, University of Amsterdam. [Google Scholar]Hochstenbach, Cody, & van Gent, Wouter (2015) An anatomy of gentrification processes: Variegating causes of neighbourhood change. Environment and Planning A, 47(7), 1480–1501. [Crossref], [Web of Science ®], [Google Scholar]Jackson, Emma, & Benson, Michaela (2014) Neither ‘Deepest, Darkest Peckham’ Nor ‘Run-of-the-Mill’ East Dulwich: The Middle Classes and their ‘Others’ in an Inner-London Neighborhood. International Journal of Urban and Regional Research, 38(4), 1195–1210. [Crossref], [Web of Science ®], [Google Scholar]Lees, Loretta (2000) A reappraisal of gentrification: Towards a ‘geography of gentrification’. Progress in Human Geography, 24(3), 389–408. [Crossref], [Web of Science ®], [Google Scholar]Lees, Loretta (2014) The urban injustices of new Labour’s “New Urban Renewal”: The case of the Aylesbury Estate in London. Antipode, 46(4), 921–947. [Crossref], [Web of Science ®], [Google Scholar]Ley, David, & Dobson, Cory (2008) Are there limits to gentrification? The contexts of impeded gentrification in Vancouver. Urban Studies, 45(12), 2471–2498. [Crossref], [Web of Science ®], [Google Scholar]McKinnish, Terra, Walsh, Randall, & White, Kirk (2010) Who Gentrifies Low-Income Neighborhoods? Journal of Urban Economics, 67(2), 180–193. [Crossref], [PubMed], [Web of Science ®], [Google Scholar]Moore, Kesha. S. (2009). Gentrification in Black face?: The return of the Black middle class to urban neighborhoods. Urban Geography, 30(2), 118–142. [Taylor & Francis Online], [Web of Science ®], [Google Scholar]ONS. (2013). A Century of Home Ownership and Renting in England and Wales. Available at: http://www.ons.gov.uk/ons/rel/census/2011-census-analysis/a-century-of-home-ownership-and-renting-in-england-and-wales/short-story-on-housing.html (accessed 16 April 2019). [Google Scholar]Paccoud, Antoine (2017) Buy-to-let gentrification: Extending social change through tenure shifts. Environment and Planning A, 49(4), 839–856. [Crossref], [Google Scholar]Paccoud, Antoine, & Mace, Alan (2018) Tenure change in London’s suburbs: Spreading gentrification or suburban upscaling?. Urban Studies, 55(6), 1313–1328. [Crossref], [Web of Science ®], [Google Scholar]Pattillo, Mary (2003) Negotiating blackness, for richer or for poorer. Ethnography, 4(1), 61–93. [Crossref], [Google Scholar]Phillips, Deborah, Davis, Cathy, & Ratcliffe, Peter (2007) British Asian narratives of urban space. Transactions of the Institute of British Geographers, 32(2), 217–234. [Crossref], [Web of Science ®], [Google Scholar]Ronald, Richard, & Kadi, Justin (2018). The revival of private landlords in Britain’s post-homeownership society. New Political Economy, 23(6),786–803. [Taylor & Francis Online], [Web of Science ®], [Google Scholar]Rose, Damaris (1984) Rethinking gentrification: Beyond the uneven development of Marxist urban theory. Environment and Planning D, 2(1),47–74. [Crossref], [Web of Science ®], [Google Scholar]Smith, Neil (1979) Toward a theory of gentrification: A back to the city movement by capital not people. Journal of the American Planning Association, 45: 538–548. [Taylor & Francis Online], [Web of Science ®], [Google Scholar]Smith, Neil (1987) Of yuppies and housing: Gentrification, social restructuring, and the urban dream. Environment and Planning D, 5(2),151–172. [Crossref], [Web of Science ®], [Google Scholar]Teresa, Benjamin F. (2016) Managing fictitious capital: The legal geography of investment and political struggle in rental housing in New York City. Environment and Planning A, 48(3),465–484. [Crossref], [Web of Science ®], [Google Scholar]Timberlake, Jeffrey M., & Johns-Wolfe, Elaina (2017) Neighborhood ethnoracial composition and gentrification in Chicago and New York, 1980 to 2010. Urban Affairs Review, 53(2), 236–272. [Crossref], [Web of Science ®], [Google Scholar]Van Criekingen, Mathieu (2009) Moving in/out of Brussels’ historical core in the early 2000s: Migration and the effects of gentrification. Urban Studies, 46(4),825–848. [Crossref], [Web of Science ®], [Google Scholar]Wachsmuth, David, & Weisler, Alexander (2018) Airbnb and the rent gap: Gentrification through the sharing economy. Environment and Planning A, 50(6),1147–1170. [Crossref], [Web of Science ®], [Google Scholar]Wacquant, Loïc, Tom Slater, & Virgílio Borges Pereira (2014) Territorial stigmatization in action. Environment and Planning A 46(6),1270–1280. [Crossref], [Web of Science ®], [Google Scholar]Walks, Alan, & August, Martine (2008) The factors inhibiting gentrification in areas with little not-market housing: Policy lessons from the Toronto experience. Urban Studies, 45(12), 2594–2625. [Crossref], [Web of Science ®], [Google Scholar]Walks, Alan, & Maaranen, Richard (2008) Gentrification, social mix, and social polarization: Testing the linkages in large Canadian cities. Urban Geography, 29(4), 293–326. [Taylor & Francis Online], [Web of Science ®], [Google Scholar]Watt, Paul (2009) Housing stock transfers, regeneration and state-led gentrification in London. Urban Policy and Research, 27(3), 229–242. [Taylor & Francis Online], [Web of Science ®], [Google Scholar]Wyly, Elvin K., & Hammel, Daniel J. (2004). Gentrification, segregation, and discrimination in the American urban system. Environment and Planning A, 36(7), 1215–1241. [Crossref], [Web of Science ®], [Google Scholar]