In many cities, there is residential segregation: people live in areas predominantly populated by members of their socially salient group. Common patterns include the residential geographies of race and class. This segregation often tracks socio-economic inequalities: those in some areas are disadvantaged with respect to the available opportunities, job prospects, wealth, education, healthcare, recreational amenities and local crime rates. It is widely agreed that we should reduce these inequalities, and this raises important questions about the geography: does seeking to address inequality give reason to support residential demographic change and, if so, of what kind?
To outline some answers, let us consider three possible demographic trajectories. One involves a wave of disadvantaged individuals moving to advantaged neighbourhoods. Let us call this disadvantaged relocation. A second involves a wave of advantaged individuals moving to disadvantaged areas. Let us call this advantaged relocation. A third is where the existing division remains broadly the same. These general trends could align with various policies. For example, states might subsidise the disadvantaged securing housing in more expensive areas, encourage the advantaged to move to disadvantaged neighbourhoods by permitting the development of modern-design flat blocks there or not target residential change at all.
In the literature, there are defences of the first and third of these possibilities. New integrationists advocate facilitating disadvantaged relocation, holding that this helps address socio-economic inequalities (see Adams Reference Adams2006; Anderson Reference Anderson2010, 112–34). Meanwhile, egalitarian pluralists argue that we should not seek to alter residential geography, but target the inequalities with community regeneration programmes (see Shelby Reference Shelby2016, 49–79; Valls Reference Valls2018, 127–50). Discussion of such views tends not to consider advantaged relocation, except occasionally to identify it as a way that community regeneration programmes might go wrong (see Madva Reference Madva2020, 223).
This is consistent with the tenor in another area of literature concerning how residential geography correlates with a further factor: cultural context. Different neighbourhoods typically have distinct customs and community activities that make up the social environment in which residents participate and their lives unfold. Critics argue that advantaged relocation displaces local residents and leads to the cultural transformation of these neighbourhoods, and that we should prevent or limit such ‘gentrification’ (see Huber and Wolkenstein Reference Huber and Wolkenstein2018; Wells Reference Wells2022). In short, two swathes of literature treat this demographic trajectory as either not worth attention or worth attention only to unpack some wrongs of residential change.
In this article, we aim to nuance this landscape and sharpen the framework for theorising residential geography by defending a certain kind of advantaged relocation. To be clear, we do not endorse demographic shifts that displace disadvantaged individuals or destroy their culture. Our point is that there is space to advance the precision in thinking about which kinds of demographic change appropriately uphold concerns with socio-economic inequalities and cultural context, and that some instances of advantaged relocation offer a surprisingly good demonstration of this. We make our case by arguing for the importance of integration in reducing socio-economic inequality through drawing attention to the effect of geographical space on prejudice, behaviour and politics, and then articulating two principles that should constrain demographic change: the first highlights the importance of distributing burdens fairly; and the second holds that transitions towards justice must avoid pressing the culture of the advantaged on to the disadvantaged. What we contend is that some instances of advantaged relocation uphold these principles, and in these cases, it is appropriately sensitive to both socio-economic inequality and cultural context.
This argument carves a distinct path through existing literature. It highlights that violation of concerns about burden distribution and cultural imposition give some reason to reject new integration programmes, but it suggests that egalitarian pluralists are too swift in deploying such objections to reject residential integration more generally. Meanwhile, our emphasis on prejudice reduction separates our account from some other attempts to defend advantaged relocation on grounds that it transfers economic and human capital to the disadvantaged. This is important because, alongside our constraining principles, it helps demarcate the instances of advantaged relocation that avoid objections to demographic change posed by egalitarian pluralists and critics of gentrification. Altogether, we steer the literature towards an avenue for integration that warrants greater attention and reappraisal, and we offer a novel theoretical approach in thinking about residential geography that has implications also for research on gentrification, housing policy and territorial occupancy.
We proceed as follows. The second section outlines a process of advantaged relocation. The third to sixth sections develop our case in its support. The seventh to tenth sections refine this argument by considering objections and critical questions. Finally, the eleventh section concludes.
An Initial Sketch
We begin by sketching the kind of advantaged relocation we aim to defend, which we call revisionist integration. This provides an overview and allows us to draw attention to certain features to serve as a guide for the rest of the article.
In revisionist integration, (1) some individuals who reside in advantaged locations move to a disadvantaged neighbourhood. This (2) does not alter the rate of outward migration from the disadvantaged area, and (3) after the initial wave, the inward movement plateaus such that there is a balanced mix of residents that remains stable over time. The result is (4) a residentially integrated neighbourhood.
As the advantaged individuals become familiar with people in their new location, (5) many of them come to hold less prejudiced attitudes towards the disadvantaged. A consequence is that (6) they tend to treat members of this group better and more commonly support social movements and political figures who advocate policies that seek to rectify the injustices that this group suffers. Meanwhile, (7) some new business and leisure facilities open to cater to the tastes of the incoming individuals, but (8) this does not increase the closure rate of existing outlets, and most make good business from both the original residents and newcomers.
This is only a partial account of revisionist integration. It focuses on aspects of the demographic trajectory and sets aside relevant features of surrounding urban policies and the social environment. We do this for the moment because we think even the numerated points may prompt some intuitive scepticism about the merits and realism of revisionist integration. Accordingly, we focus first on their normative and empirical plausibility, and then turn to showing the traction as we fill out the institutional context.
We start with point (4) of revisionist integration: a residentially integrated neighbourhood. What is the significance of this feature? Our principal rationale for supporting residential integration is its impact on (5) people's negative dispositions towards different groups and (6) how they treat others, both in interactions and through the political system. In short, its significance lies in its bearing on prejudice.
A first connection to consider is an important relationship between (5) dispositions and (6) how people act and engage politically.Footnote 1 People tend to hold prejudices about individuals they perceive as part of an out-group, and this affects behaviour. For instance, many wealthy, white individuals have beliefs that those who are poor and those who are black are less suited to positions of leadership and responsibility, attitudes that these disadvantaged individuals are to blame for their plights because they are lazy or incompetent, and feelings of contempt, distrust, and pity towards them. The advantaged who hold these dispositions are less willing to hire disadvantaged individuals, more willing to discriminate against or be violent towards them, and tend to prefer that inequalities between the groups remain or increase. These advantaged individuals also hold little political motivation to shape public policy to improve the circumstances of the disadvantaged and, indeed, align their voting behaviour to accord with their negative views about these groups. Through these mechanisms, people's prejudices have a significant effect on socio-economic inequality.
This is relevant for our purposes if residential demographics bear on the dispositions people hold and how they act. In the literature, there is some debate about whether they are related through the medium of social contact. On the one hand, new integrationists often stress that segregation leads to a lack of social bonds between those living in different areas (see Anderson Reference Anderson2010, 23–43). This is important because people tend to form prejudices towards those with whom they do not interact in certain ways, and many scenarios, from friendship circles to education and workplaces, indicate that sustained interaction improves inter-group relations and reduces prejudice (Pettigrew and Tropp Reference Pettigrew and Tropp2006). On the other hand, there is scepticism among egalitarian pluralists about the idea that residential integration will generate meaningful social bonds (see Shelby Reference Shelby2016, 70). One basis for this is that several conditions seem necessary for contact to have beneficial effects, including equal status and shared projects, and these might not arise merely from living nearby.
We will not pursue questions of social contact. This is because we think there is significance to the relationship between demographics, dispositions, and actions irrespective of whether being neighbours generates social bonds. If new integrationists are right about the potential positive impact here, our case supplements theirs. However, even if sceptics are right to doubt this, we think they overlook that demographics are important in another way.
Specifically, we think there is a persuasive case for (4) residential integration that is grounded in Ryan Enos's (Reference Enos2017, 51–78) work on how geographical space affects (5) dispositions and (6) interactions and politics. There are subtleties to this, but the basic process works as follows. As noted earlier, we tend to associate groups of individuals with certain traits, viewing and treating our in-group more favourably than our out-group. We tend to place people in these groups when they share socially salient characteristics. In societies where wealth and race have significant social bearing, they provide readily ‘accessible’ categories by which we group people as either similar or different to ourselves. Importantly, when these groups live in distinct geographical spaces, it increases the perception of them as cohesive units – more similar to those nearby and more different to those elsewhere, or, in other words, there is a sharper ‘comparative fit’. It heightens the sense of in-group and out-group. Therefore, when there is residential segregation between a wealthy, white group and a poor, black group, the former will view the latter more negatively and treat them worse than nearby peers in both interactions and politics. This effect holds independent of social contact between the groups (Enos and Celaya Reference Enos and Celaya2018).
It adds to these considerations that there is increasing evidence that residential integration has a positive impact on perceptions and action. Research finds that projects advancing residential integration have reduced group prejudices and, more generally, that it is a likely causal variable in explaining why residents in integrated neighbourhoods hold lower levels of out-group prejudice than those in segregated areas (Barnhardt Reference Barnhardt2010; Laurence et al. 2019; Oliver and Wong Reference Oliver and Wong2003).Footnote 2 There is also evidence that inter-group mistreatment reduces over time in residential integration projects and that level of neighbourhood integration has a notable impact on commitment and support among the advantaged for addressing the injustices the disadvantaged face (Brown et al. 2021; Kinder and Mendelberg Reference Kinder and Mendelberg1995; Patterson Reference Patterson1998, 191). It is undoubtedly wise to exercise caution here. Further research might refine our understanding of this connection and its mechanisms. However, we think it is a sufficient basis for a presumption in favour of pursuing residential integration on grounds that it tackles the way segregated geographical space affects prejudices. Based on the evidence already identified about the effects of these attitudes, this indicates that we have reason to advance (4) residential integration on grounds that, independent of any bearing on social contact, its realignment of geographical space positively impacts (5) dispositions and (6) interactions and political engagement in a way that will tackle socio-economic inequality.
The Principles of Integration
Having set out our case for (4) residential integration, we now turn to the idea of advancing towards this end by (1) some individuals who reside in advantaged areas relocating to a disadvantaged neighbourhood. After all, this is not the only way to pursue integration. An alternative, advocated by new integrationists, is for the state to facilitate disadvantaged individuals relocating to advantaged neighbourhoods (Adams Reference Adams2006, 268–76; Anderson Reference Anderson2010, 118–22). What reasons are there for favouring one means of integration or another? Our answer to this question holds that there are important moral concerns about how we pursue this goal. We elaborate these through a critical analysis of new integration, before turning to the comparative strengths of revisionist integration.
To focus attention, we consider two prominent examples of new integration in US policy. One is the Gautreaux Program, which gave housing vouchers to disadvantaged black families to relocate to areas predominantly populated by wealthy white families. Another is the Moving to Opportunity Program, which moved families from high-poverty neighbourhoods to lower-poverty neighbourhoods through public housing reassignment. We noted earlier research indicating that these programmes positively impacted problematic dispositions of the advantaged. There is also evidence that disadvantaged individuals who relocated experienced increases in income, declining exposure to poverty and improvements in educational attainment (Chetty, Hendren and Katz Reference Chetty, Hendren and Katz2016; Hartman and Squires Reference Hartman and Squires2010; Rosenbaum Reference Rosenbaum, Cohen, Deckers and Fiss2003). However, there are two important problems with new integration programmes.
First, there is the problem of burden distribution. One group that suffers this problem is those who new integration asks to move from disadvantaged to advantaged neighbourhoods. Evidence shows that a notable proportion of these individuals faced considerable hostility from existing residents, particularly early on, and were subject to verbal harassment, threats, physical violence and attacks on their property (Rosenbaum et al. 1991, 454–5). Meanwhile, relocating disrupts existing social ties within old neighbourhoods, entailing the loss of various kinds of informal support (Valls Reference Valls2018, 141). To make this vivid, let us imagine a single mother who relies on her neighbour for transport around town and relies on her uncle to look after her children while she is at work. Even if we can guarantee that this individual's networks will eventually be replaced by ‘better’ opportunities in an advantaged neighbourhood, losing her existing ties during the transition is clearly a huge cost.
We should also consider those who remain in disadvantaged neighbourhoods. Whatever socio-economic benefits result for those who move, these benefits do not readily accrue to the neighbourhoods they leave (Valls Reference Valls2018, 141–2). Meanwhile, several of the costs of relocation also affect those who remain. They too lose neighbours and family who help with transport and childcare. In the same way, these programmes deplete the pool of human capital surrounding them and their children (Brooks Reference Brooks1996, 55).
In contrast, new integration programmes ask advantaged individuals to bear few of such costs (Valls Reference Valls2010, 474). They may provide some support through their taxes. However, they do not incur the risks of relocation or dwindling opportunity in their neighbourhoods. Indeed, they can continue their lives much as before, in the same house near the same family and friends. This is one reason why such programmes attract the complaint that they are ‘a one-way street’ (Ture and Hamilton Reference Ture and Hamilton1992, 54–5).
Altogether, new integration distributes the costs of addressing socio-economic inequality in the wrong way, loading the burdens disproportionately on the already disadvantaged. We think these problems point towards a principle of the following kind:
Principle of Transitional Burden Distribution: All else equal, the disadvantaged should not bear the costs of moving towards a more just society.Footnote 3
The violation of this principle is one way that new integration goes wrong.
Another issue we must take into account in residential integration is what we can call the problem of cultural assimilation. We noted earlier that neighbourhoods often have distinct customs and community activities. A problem with new integration is how it treats this cultural division. These programmes tackle socio-economic inequality in a way that asks those who suffer injustice to assimilate to the neighbourhood cultural context of those who are wrongfully privileged and often complicit in this injustice. Stated at its most troubling, Kwame Ture and Charles V. Hamilton argue that new integration programmes make access to fair opportunities dependent on the demand that ‘black people must give up their identity, deny their heritage’ and conform to the culture of those who have oppressed them (Ture and Hamilton Reference Ture and Hamilton1992, 54–5; see also Brooks Reference Brooks1996, 56–7). Being asked to give up their background and its place in collective culture stands in tension with the respect to which the disadvantaged are entitled and is potentially alienating (Boxill Reference Boxill1992).
In making such a demand, these programmes also cast a negative, perhaps even stigmatising, judgement on disadvantaged neighbourhoods. As Mary Pattillo (Reference Pattillo2014) puts it, new integration ‘dwells on … the relatively problematic nature of Black people and Black spaces’. It has some similarities to the argument that it is a ‘culture of poverty’ in these neighbourhoods that is the central cause of disadvantage and that the solution is to stamp out this self-destructive pathology and replace it with the culture of the advantaged. Ture and Hamilton (Reference Ture and Hamilton1992, 54–5) argue that, in this way, such programmes feed historical hierarchies, perpetuating ‘the idea that “white” is automatically superior and “black” is by definition inferior’. Meanwhile, akin to a concern that we raise about burden distribution, it appears to locate the moral problems the wrong way around. It misplaces blame for injustice with the disadvantaged, rather than with those or the system that perpetuate it.
Defenders of new integration often respond to such concerns by highlighting that cultural division tends to lessen over time as the disadvantaged who relocate acclimatise to their surroundings. As Elizabeth Anderson (Reference Anderson2010, 118–19) argues in relation to the Gautreaux Program, although participants initially felt uncomfortable with some social customs in their new neighbourhoods, they reported many benefits ‘once they adjusted to the … norms’. However, this captures precisely the problem of cultural assimilation. While the programmes may have some positive outcomes, they demand residential integration on the terms of the advantaged.
Based on these concerns, we think that it is valuable to supplement the principle of transitional burden distribution with the following:
Principle of Transitional Cultural Respect: All else equal, the transition towards a more just society should not involve asking the disadvantaged to assimilate to the culture of the advantaged or convey that it is morally superior.
We think this principle is intuitive and the concerns raised by new integration support it. Plausibly, it has limits. Tommie Shelby (Reference Shelby2005, 180) expresses the worry that certain aspects of culture in disadvantaged neighbourhoods might be ‘obnoxious, immoral, pathological, or otherwise unappealing’, and holds that we should not embrace these elements. The principle of transitional cultural respect is compatible with this, and perhaps a refined version of the principle needs to draw an appropriate line. However, this is not crucial for our purposes. What matters here is that many aspects of culture among disadvantaged individuals are not objectionable, and our concern with new integration is that it requires assimilation that imposes upon even these elements. Our principle states that processes of residential integration should avoid this.
In sum, our objection to new integration is that it transgresses two important parameters of moving towards a more just society. It violates a concern for burden distribution and cultural respect in transition. This, we think, counts against advancing residential integration through such policies, at least if there is an alternative that would better conform to these principles. This brings us back to revisionist integration.
Integration on Fair Terms
While we might advance residential integration with advantaged relocation or disadvantaged relocation, our claim is that revisionist integration better upholds the principles of transitional burden distribution and cultural respect. We demonstrate this by returning to certain other features of this scenario.
First, let us recall that our sketch involved (1) a wave of advantaged individuals relocating to disadvantaged areas that (2) does not alter the rate of outward migration from the disadvantaged area, and after a while, (3) the inward movement plateaus such that there is a balanced mix of residents that remains stable over time. Meanwhile, (7) some new business and leisure facilities open to cater to the tastes of the incoming individuals, but (8) this does not increase the closure rate of existing outlets, and most make good business from both original residents and newcomers.
These features ensure that the process conforms to the concerns motivating the principle of transitional burden distribution. Since the integration occurs within disadvantaged areas and it places no additional pressure on people or the amenities there to relocate, residents will not lose their social ties or the human capital of their long-term neighbours. Thus, it does not ask the disadvantaged to bear a notable burden of the integration process.
Similarly, these features indicate that revisionist integration can uphold the principle of transitional cultural respect. Although Point 7 involves some elements of change, Point 8 means that the transition is not one that asks for assimilation to the culture of the advantaged. That the original residents and amenities remain stable means that the neighbourhood retains some of its distinctive cultural context, and there is a familiar social environment in which lives unfold. In essence, there is some retention and some fusion, but not the imposition of the culture of the advantaged on the disadvantaged.
It adds to these considerations to recall the specific connection that we emphasised between revisionist integration and socio-economic inequality. Specifically, we centred attention on advancing residential integration (5) to reduce the prejudices held by the advantaged and (6) to improve how they interact with the disadvantaged and engage politically in support of addressing the injustices that this group suffers. Rather than implying that there is a superior culture to which the disadvantaged should assimilate, this locates the problem in the attitudes and actions of the advantaged and proposes respectful engagement with the people and places that suffer injustice as the solution. In short, it proposes residential integration that is fair to the disadvantaged and takes place on terms respectful to them.
Revisionist Integration and Egalitarian Pluralism
At this point, it is useful to draw together how our arguments intersect with those of egalitarian pluralists. We do this by considering a different concern they raise about residential integration regarding its fittingness as a response to injustice (see Shelby Reference Shelby2016, 77–8; Valls Reference Valls2018, 146–8). For all we have said, they might ask: if the aim is to address socio-economic inequality without violating relevant principles of transition, why not simply redistribute wealth? Given the end, residential integration might seem an unnecessarily roundabout means. It may also appear ill-fitting when we see socio-economic inequality as a wrong done by the advantaged to the disadvantaged that warrants rectification. The right response to such scenarios, the argument continues, is not circuitous detours, but to pay what is owed.
We do not think there is much disagreement between our view and that of egalitarian pluralists here. We concur that addressing the injustices of inequality readily speaks in favour of redistribution. Our aim has been to explore what this plausibly entails, indicating reasons to think residential integration's impact on prejudice has an important role in generating the relevant political motivation and action. To put this concretely, let us consider a redistributive policy proposed by egalitarian pluralists: taxation to fund community regeneration programmes. What motivates our view is that taxation is a political enterprise, one that relies on attitudes and behaviour of the populace, particularly at the voting booth. Our concern is that such policies are difficult to advance, let alone maintain, in the face of the prejudice caused by the geographical space between different groups. Accordingly, what we have sought to express is not a contestation with egalitarian pluralists on the importance of redistribution as a means to tackle socio-economic inequality. Rather, we see our proposal as complementary, suggesting that residential integration is part of the suite of actions useful to this effect.
Even showing that revisionist integration has value in this respect may fail to persuade if it incurs the problems of burden distribution and cultural assimilation. On this as well, we share the concerns that egalitarian pluralists press against new integration and agree that they are weighty enough to caution against this policy even if it has some benefits. Here, our argument has been that problems with new integration are not problems with residential integration per se and, specifically, that they do not arise in revisionist integration. Accordingly, such concerns do not give reason to remove this option from our strategy suite, and our intersection with egalitarian pluralists extends to elaborating a way to advance redistribution that is compatible with normative parameters they think should constrain the pursuit of this aim.
From this juncture, we think it will take our case forward best to consider certain objections. This is because some may doubt whether our arguments thus far have done enough to establish the credentials that we have attributed to revisionist integration. One worry that might appear to loom large is that, despite our claims, revisionist integration does not avoid the problems of new integration. Critics often charge advantaged relocation with concerns about burden distribution and cultural assimilation, arguing that they are hallmark features of ‘gentrification’. Thus, it is useful to turn next to the fear that our account of revisionist integration is quixotic, not resembling how such processes happen in the real world. Then, we discuss the charge that it might harbour other problematic features of gentrification. Subsequently, we consider whether revisionist integration violates a permission for the disadvantaged to self-segregate. Finally, we explore questions about there being a viable institutional context in which it could occur.
Let us begin with a definition of gentrification. Jakob Huber and Fabio Wolkenstein (Reference Huber and Wolkenstein2018, 380–1) describe the process as involving four aspects. It involves social change, where middle-class demand for housing in disadvantaged neighbourhoods leads to increases in rents and forces existing residents to relocate. It involves spatial change, where more affordable housing is replaced by luxury flat blocks and local cafes are replaced by artisan coffee shops. It involves transformational change, where new languages, social practices, commercial services and community activities reshape the ‘lifeworld’ of the area such that long-term inhabitants cannot sustain their former cultural context. At the furthest end is expulsion change, where these other effects lead to the departure of the remaining original residents.
Processes of these kinds are widely considered problematic, and, for what it is worth, we think that the principles outlined earlier provide some account of their wrongs. The displacement of long-term residents places costs on them likely to violate the principle of transitional burden distribution. Meanwhile, the transformational change of the cultural context in a disadvantaged neighbourhood into a social environment responsive to the social practices of the wealthy is likely to violate the principle of transitional cultural respect's concern with assimilation. Indeed, when there is expulsion change, it does not even advance residential integration. Thus, we hope that it is clear that our arguments do not support – indeed, plausibly condemn – processes involving these effects.
This marks a useful point to consider another argument occasionally made for advantaged relocation. This is the claim that it involves wealthy individuals bringing financial and human capital that, as advocates sometimes put it, ‘rubs off’ on disadvantaged neighbourhoods and those who live there, swelling the tax base and leading to increases in job prospects and educational attainment, and decreases in poverty exposure and crime (Byrne Reference Byrne2002, 415–24; Duany Reference Duany2001, 37). We also see some of these effects positively – indeed, increased tax returns that redistribute wealth, in particular, is a measure for which we have already indicated support – and might agree that they are a valuable by-product of advantaged relocation. However, we think it is important to exercise caution with deploying this kind of argument as its principal rationale. This is for reasons explored in our discussion of new integration, namely, that placing emphasis on how privilege ‘rubs off’ on the disadvantaged harbours a problematic message about moral superiority of the advantaged and their culture. In similar vein, we are wary of this argument's potential to support instances of gentrification that, in the ways we indicate in the previous paragraph, violate our principles of burden distribution and cultural respect.
However, for our purposes, it is important to highlight that such concerns do not arise in all versions of advantaged relocation. Specifically, among the main features of revisionist integration is that it does not displace original residents or emphasise transforming the context to the culture of the advantaged. In this respect, it does not involve the social, spatial, transformational and expulsion change of gentrification.Footnote 4 Put another way, we should not treat advantaged relocation as synonymous with gentrification. As revisionist integration shows, some scenarios that fit into the former category do not fit into the latter category.
It is here that critics are likely to press the quixotic objection, arguing that whatever the technical possibility of separating these categories, it is a distinction without substantive bearing. They may complain that advantaged relocation (almost) always does constitute gentrification. In other words, the set of scenarios where advantaged individuals move to disadvantaged neighbourhoods that does not involve social, spatial, transformational and expulsion change is, in reality, an empty set.
We have two responses to this. The first is empirical and holds that revisionist integration is not quixotic. We think that the plausibility of this reply depends most on the evidence relating to three features of the scenario. One is the claim that the relocation of advantaged individuals (2) does not alter the rate of outward migration from disadvantaged areas. Another is the claim that (3) the inward movement plateaus such that there is a balanced mix of residents that remains stable over time. The last is that (8) it does not increase the closure rate of existing outlets, and most make good business from both original residents and newcomers.
While research into advantaged relocation is still developing, there is emerging evidence from North America and the UK that these features are reasonably common. On Point 2, advantaged relocation often has little effect on the out-migration rate of disadvantaged individuals or in terms of raising their rents (Asquith, Mast and Reed Reference Asquith, Mast and Reed2021; Ding, Hwang and Divringi Reference Ding, Hwang and Divringi2016; Dragen, Ellen and Glied Reference Dragen, Ellen and Glied2019; Freeman, Cassola and Cai Reference Freeman, Cassola and Cai2016; McKinnish, Walsh and White Reference McKinnish, Walsh and White2010). This is largely due to one way in which it often differs from the description of gentrification outlined earlier. In Huber and Wolkenstein's account, social change involves middle-class demand for existing housing pushing up rents, while spatial change involves the replacement of affordable accommodation with pricier options. However, rather than moving into or supplanting accommodation formerly occupied by the disadvantaged, it tends to be new accommodation that attracts the advantaged, typically modern flat blocks built in unused spaces, such as in place of derelict warehouses. In other words, the increase in demand matches an increase in supply and functions more in a parallel market rather than bearing on the accommodation of the disadvantaged. On Point 3, comprehensive longitudinal studies of neighbourhoods that experienced waves of advantaged relocation find that many see the inward migration plateau over time, resulting in a stable balance of original and incoming residents (Ellen and Torrats-Espinosa Reference Ellen and Torrats-Espinosa2019; Walks and Maaranen Reference Walks and Maaranen2008). Quite plausibly as a result of the demographic balance that develops, there is evidence on Point 8 that the arrival of wealthy individuals into disadvantaged neighbourhoods tends not to change the rate at which local businesses close and move out, and that it has benefits for long-standing enterprises (Glaeser, Kim and Luca Reference Glaeser, Kim and Luca2018; Meltzer Reference Meltzer2016).
We think this evidence supports the nuance we propose in debates about residential integration. We should not overstate its significance. We could easily identify cases where advantaged relocation has taken the form of gentrification, and the longitudinal studies we cite earlier do find instances in which this occurs and express caution about potential trends in this direction. However, neither should we overstate the significance of these cases. Ingrid Gould Ellen and Gerald Torrats-Espinosa make this point aptly in their analysis. They highlight that we should be cognisant of cases like Williamsburg, Brooklyn, where the demographics changed from a predominantly disadvantaged, black and Hispanic community to a wealthy, white community over a period of roughly 25 years from the early 1990s (Ellen and Torrats-Espinosa Reference Ellen and Torrats-Espinosa2019, 846–7). However, they note, ‘such dramatic … change is unusual’: only 0.4 per cent of the New York neighbourhoods they studied that saw advantaged relocation in the same period experienced changes of the kind in Williamsburg (Ellen and Torrats-Espinosa Reference Ellen and Torrats-Espinosa2019, 847). Indeed, of the 930 neighbourhoods predominantly populated by disadvantaged groups that saw advantaged relocation in the 1980s and 1990s altogether, the number that remained so by 2016 was 746 (80 per cent) and the number that became predominantly populated by advantaged individuals was only eight (0.8 per cent) (Ellen and Torrats-Espinosa Reference Ellen and Torrats-Espinosa2019, 845). Simply put, the category of advantaged relocation that is not gentrification is not an (almost) empty set. Instances with the features of revisionist integration occur often enough to deny that they are quixotic.
Our second reply adds to this in terms of plausible possibilities. The thought here is that we can consider ways in which states might intervene in advantaged relocation to ensure that it falls short of gentrification and avoids violating the principles of transitional burden distribution and cultural respect. We do not think that this point will surprise or trouble critics of gentrification because they too argue that regulation is a reasonable solution to their concerns (see Huber and Wolkenstein Reference Huber and Wolkenstein2018, 386–8). Common proposals are rent control and long-term leases. Another is Margaret Kohn's (Reference Kohn, Bell and Zacha2020, 193) argument for maintaining affordable flats that are accessible to the disadvantaged. However, our concern at this point is not what constitutes the best policy choices (we return to such matters later). What is significant for our purposes is that there are some practical options. Again, this suggests that it is too quick to dismiss revisionist integration as quixotic.
Even if revisionist integration falls short of gentrification, critics may fear that it is wrongful in similar ways. We have argued that it avoids concerns about burden distribution and cultural assimilation. However, there are other claims about the wrongs of gentrification that might pose a threat to our view. Perhaps most notably, one argument in recent literature is that gentrification denies people their ‘occupancy rights’ (Huber and Wolkenstein Reference Huber and Wolkenstein2018). Does revisionist integration violate this concern?
The idea of an occupancy right is that people have a weighty interest in being able to pursue their projects in a particular location (Stilz Reference Stilz2013, 324–56). We have both autonomy and well-being interests in forming and following a life plan. An important aspect of this is that life plans are located, in the sense that we develop them somewhere. We shape and pursue them within a specific place and cultural context, and disruption to our spatial circumstances interferes with this. This might be thought to generate a right to remain within the location – the physical space and context – where we develop our plans, a right that grants immunity from expulsion and devastation of our social environment.
Whether occupancy rights pose a challenge to our view depends on their shape and weight in different cases. One kind of scenario is where there is complete dissociation or devastation of a locale. This is the situation in one of the main cases that motivates arguments for occupancy rights: where colonisers drove Indigenous peoples from their lands. Here, the idea that there is a violation of a strong occupancy right looks readily plausible. Perhaps the same is true if gentrification involves complete location changes too. However, for reasons already developed, this is consistent with our defence of revisionist integration both empirically and theoretically. As we described earlier, many instances of advantaged relocation do not increase the rate of outward migration or completely transform the local culture, and the principles that underlie revisionist integration demand that we regulate to prevent such outcomes.
It is worth pausing here to consider a subtlety. To say that revisionist integration does not increase outward migration is not to say that there is no outward migration from disadvantaged areas. Nor, therefore, is it to say that residents in these neighbourhoods do not suffer displacement. It is plausible that many do and that this violates their occupancy rights. The relevance of some forms of advantaged relocation not increasing outward migration is that we should attribute these rights violations to other causes, such as socio-economic inequality. We should not charge revisionist integration with creating these wrongs through displacement.
The more interesting scenario for our purposes is when neighbourhoods see the arrival of new residents and partial cultural change. The notion of occupancy rights still has some force here. In the case of Indigenous peoples, it is plausible that even a partial overtake of the space and culture is problematic. However, it is worth considering our reaction to other situations. One is when some disadvantaged individuals take up public housing in an advantaged neighbourhood and some new amenities open to cater to this incoming group. Another is where international immigration supplements the population in a disadvantaged area and this brings new businesses and leisure facilities, such as alternative places of worship, alongside the long-term amenities. In these cases, we do not have the intuition that there is a problematic violation of an occupancy right.
Two factors might be involved in generating this result. One thing we think differentiates the case of Indigenous peoples from these other scenarios lies in the tightness of fit between life plans and locations. In the case of Indigenous peoples, there is a sense in which life plans are so closely intertwined with the location that even partial changes might be notably disruptive. In contrast, the change in the culture of the advantaged neighbourhood is not of the kind that poses a serious obstacle to residents accessing the space and context that their projects require. Similarly, that immigration adds people and amenities to a disadvantaged area does not seem sufficient for original residents to claim that their life plans are culturally unavailable, or even objectionably offset. Another pertinent factor is how we judge the actions or interests of others. We give no positive moral value to colonial expansion. In contrast, we see it as good when disadvantaged individuals and those who immigrate to a new state find valuable social, political and economic opportunities. These factors might suggest that partial demographic and cultural changes do not pose a threat to occupancy rights when they do not interfere greatly with original residents' life plans or that certain other factors might outweigh any such threat.
This is important for our purposes because arguments of both kinds are available for revisionist integration and, indeed, appear consistent with what critics of gentrification and advocates of occupancy rights argue. It might be that occupancy rights do not have the relevant tightness of fit to reject partial cultural change in this case. Huber and Wolkenstein (Reference Huber and Wolkenstein2018, 386) seem to take this line when they argue that their case for occupancy rights does not mean ‘neighbourhoods should be hermetically sealed from newcomers and their transformative impact’. A plausible basis for this position is that, like the cases of disadvantaged relocation and immigration described earlier, revisionist integration does not present much obstacle to residents of disadvantaged neighbourhoods pursuing their existing life plans. Perhaps there is no occupancy right violation involved here.
Another possibility is that if there is a violation, other considerations outweigh it. Anna Stilz (Reference Stilz2013, 343–4) highlights this prospect in her account of occupancy rights. She outlines a scenario where a group of individuals who have shaped their projects within a tight-knit mining community face needing to relocate when business becomes unprofitable. She argues that despite their occupancy rights, they are not entitled to subsidies or protection against the decline of the industry because there is the competing interest in a dynamic market economy.Footnote 5 This is relevant to revisionist integration because it too involves an important benefit: the merits of residential integration detailed earlier.
We need not take a stand on which of these points is more pertinent to revisionist integration. What is key for our argument is that accounts of occupancy rights and their relevance to gentrification align with what follows from our principles of transitional burden distribution and cultural respect. When demographic changes displace disadvantaged individuals and completely overhaul existing neighbourhoods, we should object. However, when changes are more partial, there are scenarios where our reasons for objection dissipate, and we think revisionist integration is one such case.
There is another way that critics might pose an objection to the partial demographic and cultural changes that revisionist integration involves. This appeals to the idea that disadvantaged individuals have a permission to maintain some kind of separation from the advantaged in their society. That it transgresses this permission is another complaint critics often lodge against new integration (see Shelby Reference Shelby2016, 59–62; Young Reference Young1999, 244–5). Is a similar objection applicable to revisionist integration?
To answer this question, it is useful to distinguish between two kinds of separation. One is the idea of separatism, where the aim is strict residential division to establish spaces for the disadvantaged that are relatively shielded from the domination of the advantaged. This has some historical pedigree (Robinson Reference Robinson2001). However, there was a general shift away from endorsing separatism, even among those who initially expressed support, such as Malcolm X and Roy Innis. One reason for this is general population dispersal. For example, as a result of many black individuals in the US moving from southern rural areas to northern cities, the territorial location for separatism is difficult to identify (Valls Reference Valls2010, 473).Footnote 6 Another reason is the increasingly ubiquitous reach of economic and political power over the last century or so. Against this background, it has become unclear whether separatism can serve its intended goals – as Shelby (Reference Shelby2005, 141) puts it, even strict residential division offers ‘little escape from white influence’ in the world of twenty-first century politics.
The prominent position in contemporary writing is defences of more limited forms of separation or ‘self-segregation’ (Brooks Reference Brooks2009, 63–88; Merry Reference Merry2013, 67–92; Shelby Reference Shelby2016, 59–62).Footnote 7 The aims of self-segregation centre on creating a nurturing environment for the disadvantaged to cultivate self-respect, establishing a basis for security and the development of political solidarity, preserving culture, and securing a proportional degree of political autonomy over the institutions that affect the lives of the disadvantaged. This is different to separatism, in that advocates of self-segregation are not averse to disadvantaged neighbourhoods having some population balance or to advantaged individuals participating in the cultural context and institutional context there (Brooks Reference Brooks1996, 190). The point is that disadvantaged individuals have a permission to live in communities sufficiently separated to sustain security, solidarity and mobilisation.
This is important because it means that self-segregation needs only that there is a critical mass of disadvantaged individuals in a neighbourhood. Advocates estimate that it requires making up 25–50 per cent of the local population (Shelby Reference Shelby2016, 61). However, even if we set the proportion somewhat higher, it remains compatible with our account of revisionist integration, where the incoming population plateaus and there is a balance of groups in the neighbourhood. We should exercise sensitivity as to what exactly self-segregation requires here. In our earlier commentary, we specified only that revisionist integration results in a balanced population. In light of concerns about self-segregation, we might stipulate this more precisely, holding, say, that the inward movement of the advantaged must plateau below the group constituting 50 per cent of the local population. Such a specification is consistent with where the longitudinal studies we cited earlier suggest demographics do plateau. However, again, we are open to the need for policies that regulate the process to achieve this. Likewise, it is compatible with, perhaps even indicated by the principles underpinning, revisionist integration to endorse policies required to ensure other features of a secure environment needed for solidarity and mobilisation, such as quotas for local political representation. What matters for our purposes is that refinements of this kind are readily in line with our general defence of revisionist integration, and, thus, we do not think it is incompatible with self-segregation.
The Policy Environment
It might seem that another issue has now begun to loom large: institutional context. In our overview of revisionist integration, we expressly omitted details of surrounding urban policies and social environment. However, such considerations have crept into our analysis in response to objections. It is useful now to give these matters fuller attention to answer two questions. One is: what does revisionist integration assume, involve or require in terms of state functionalities and individual motivations and behaviour (particularly in their migratory decisions)? A second is: is it plausible that they can play the role ascribed to them, or are they overly idealistic?
We can address the second of these questions while remaining somewhat open on the first. To elaborate, let us return to the concern that advantaged relocation may lead to gentrification that displaces the disadvantaged. Some may fear this could happen quite swiftly or, at least, doubt that there are ways to prevent it in the long term once we open the door. However, it is worth noting the vast array of commonplace tools at policy makers' disposal for incentivising and discouraging such migration that work with existing motivations and behaviour. A merely partial list includes rental assistance vouchers, eviction controls (with legal support), loans and tax breaks to help existing residents purchase their property, housing conversion restrictions, inclusionary zoning, and community land trusts (Been Reference Been and Herbert2018). While research is ongoing about what is most effective in shaping advantaged relocation, the historical pedigree of these tools in regulating migration patterns and demographics give some reason to believe that there is an available combination that can marshal revisionist integration as we suggest.
It adds to this to consider two connections to institutional context that emerge from our foregoing analysis. One is the evidence we cited on advantaged relocation pertaining to instances where the state permits newbuild apartment blocks. When it occurs this way, it tends to tap advantaged individuals' interest in securing accommodation that is cheaper than their current location, but it operates in a parallel housing market in the disadvantaged area. For this reason, the relocation flow reduces significantly when the new flats are taken and has little effect on existing residences or rents in the neighbourhood. Another is the longitudinal studies we discussed. The finding in this literature is that the presence and maintenance of public housing can play a pivotal role in balancing demographic ratios at the level we propose over some decades (Ellen and Torrats-Espinosa Reference Ellen and Torrats-Espinosa2019, 847). This is partly because it also works in a parallel market that is shielded from advantaged individuals seeking housing and the possible pressures on rent that could result.
Even sketching only some such institutional context for revisionist integration by drawing on our earlier analysis speaks to its plausibility. This is because this investigation worked inductively with a set of literature that has explored the dynamics and effects of how advantaged relocation has taken place in recent decades. Accordingly, the insights we have drawn about the possible policies involved are grounded in the empirical reality of the contemporary, non-ideal world. For the same reason, there is some basis to think that these measures can navigate the interests and behaviour of both advantaged and disadvantaged individuals, namely, that they appear to have done so in actual instances of revisionist integration.
This is not to claim that we have identified and irrefutably defended all relevant urban policies and aspects of social environment. Among other things, the nature of the housing market, building regulations and overall governance will matter. Meanwhile, the empirics in which we grounded our view were drawn from specific contexts, primarily North America and the UK. We must also always avoid overstating the conclusiveness of any evidence, particularly given the precise aims of revisionist integration and whether we can achieve them over more than a couple of decades.
However, it is important not to underestimate the significance of our analysis either. To see this, it is worth recalling that our investigation has tried to explore not only the normative parameters in the relationship between tackling inequality and residential integration, but also the practicalities that connect them and that are involved in taking progressive action. When engaging with these issues, difficult questions inevitably arise about accounting for the entire array of relevant factors and guarantees about their implementation and upshots. Given the possibility mentioned earlier that community regeneration programmes will lead to gentrification, egalitarian pluralists must answer the same practical questions as us about the institutional context needed to avoid this outcome. Meanwhile, new integrationists face parallel concerns about disadvantaged relocation potentially leading to ‘white flight’. There are obvious limitations to answering such questions. What it is possible to offer is a careful analysis of whether existing evidence gives a proposal some grounding and plausibility. We think that having made progress of this kind suffices to show that revisionist integration has traction in several pertinent contexts, which might generalise to various other settings that have similarities, and, if nothing else, is mileage valuable to others exploring these questions.
We have discussed the main tenets of our argument enough that the most illuminating conclusion will be to comment on two issues in debates about demographic change that we have not discussed at length. The first concerns the idea of moral imperatives. In the literature, there is some focus on showing that segregation is the ‘linchpin’ of inequality (Anderson Reference Anderson2010, 67) or, in contest, that ‘integration … is not a requirement of justice’ (Shelby Reference Shelby2016, 67). If we were to place our view in such terms, it would fall closest to Andrew Valls' (Reference Valls2018, 144) description of articulating what we might call a contingent imperative: a claim that revisionist integration plausibly has a place in the overall strategy ‘likely to be effective in achieving … equality’ given a ‘number of contingent facts’. However, we have sought to avoid this language, and in line with this, our earlier discussion of egalitarian pluralism focused on how our view relates to, rather than competes with, redistribution. Some may doubt whether this fully answers the question of its comparative imperativeness or even wonder if we have strayed in focus from questions of justice to those of political feasibility. For what it is worth, our position on this is that what tackling inequality (let alone wider matters of justice connected to it) requires is likely an array of actions working together with various elements of mutual reinforcement, making it difficult to speak of specific imperatives or pairwise comparisons. Defending this claim and investigating the available combinations is a task for another occasion. Our connected aim here – and what we can achieve in one article – has been to show only that we should not overlook revisionist integration in this exploration. It suffices for this to place it among the set that plausibly have some role in tackling inequality within the constraints set by relevant principles of transition.
In considering the options in such a set, one area we think does not give any cause for disagreement are questions about the structural nature of injustice and the agency of the oppressed. Shelby (Reference Shelby2016, 2–3) presses these points in his criticisms of new integration, holding that it places too much weight on a ‘narrowly targeted’ intervention that works within an unquestioned background structure and imposes change upon the disadvantaged in a technocratic manner rather than through their collective voice. Some argue that these are not fair criticisms of (all) new integrationists (Madva Reference Madva2020, 213–15). For our purposes, what is important is that they do not apply to our view. We hope that it is clear from our arguments that revisionist integration does not accept the status quo background structure. Meanwhile, we have not taken a position on the political system through which demographic change must occur. This flows from our view about the role of political philosophers, namely, to tender an argument for political consideration. Whether and how to take ideas forward has not been our concern. These are matters for a democratic polity that should be properly responsive to the voices of the disadvantaged.
Rather, the focus of our attention has been the merits and parameters of possible trajectories for residential demographics, setting out the credentials of one in particular that is often viewed with suspicion. While issues remain for further attention, we hope to have shown that we should not disregard possible avenues of advantaged relocation too swiftly. It is, we think, worth considering a viable form of revisionist integration that can harness the benefits of residential integration and take place on fair terms for the disadvantaged.
For discussion of earlier drafts of this article, we thank Will Abel, Rosario Aguilar, David Axelsen, Hartmut Behr, Derek Bell, Matthew Clayton, Katarina Fragoso, Signy Gutnick Allen, Johannes Kniess, Nicola Mulkeen, Tom Parr, Andrei Poama, Tommie Shelby, Kai Spiekermann, Terri-Anne Teo and Katy Wells. We are also grateful to audiences at Institute of Higher Studies in Economics and Commerce Business School Paris, University of São Paulo Law School, London School of Economics Political Theory Research Seminar Series, Punjabi University and Newcastle University's Political Theory Seminar Series. We are also very grateful for comments from Paul Bou-Habib and three anonymous referees at the British Journal of Political Science.