Although the ebb and flow of gentrification in Britain today appears to have declined, it
is important to understand what form it takes when the factors that would be thought to
halt the process have not done so. Recent observation has shown gentrification to be
occurring in London (Warde 1991, Bridge 1993a and 1993b, Lees 1994) and the northeast of England (Cameron 1992) 1 . Later interviews with estate agents confirmed that
having climbed out of the recession certain parts of the market in parts of London,
notably gentrified areas and the upper end of the market, were as buoyant as they ever
One can look for gentrification in other areas, such as the effect of the 1988 Rent Act
that has deregulated tenancies so that low income tenants are being replaced by high
income tenants as deregulation has led to insecurity of tenure (Jew, 1994). Housing
benefit changes have also pushed down the ability of tenants to pay high rents.
Lower levels of gentrification today may be due to a “gentrification saturation point”
(Atkinson, 1995). In such circumstances supplies of gentrifiable property and gentrifiers
are exhausted leading to a temporary cessation. However, even if this is the case
occupational restructuring and changes in taste, about what property may be deemed
gentrifiable, may lead to new movements being made at any time. Recent trends in the
conversion of schoolhouse, ecclesiastical and industrial property in working class areas
are examples of this redefinition of gentrifiable property and may open new avenues that
lead to the gentrification of surrounding housing.
The extent to which the ‘right to buy’ legislation has led to incumbent upgrading
through a tenure shift from public rented to owner occupied has not been speculated
upon and is clearly an issue of what constitute class and status boundaries. If class may
be defined by tenure then the right to buy legislation has created the largest single
‘journalistic accounts have been much more prolific in their identification of gentrified areas but this
may be as much the desire to create as to observe patterns of gentrification.
Gender has been argued to be an increasingly important component involved in the
gentrification of certain areas of Hackney (Warde, 1991) although the extent to which
this was an overriding concern was debated by Butler and Hamnett (1994) who argued
that class was still the dominant characteristic although the increasing market power of
female headed households was becoming more significant. This clearly needs more
analysis particularly as women become a larger and better paid proportion of the
professional and managerial groups.
If Castells is correct in asserting the major cleavage in class relations to be based around
access and non-access to information (Castells, 1994) this may mean that in
combination with a particular taste in housing the ‘open-collar’ worker may form a new
gentrifying group occupying new areas in ‘electronic cottages’ located in any area due to
the ability of information technology to be located anywhere. In view of what has
already been argued is it possible that gentrifiers as an identifiable group are a highly
fragmented and heterogeneous set that require more sophisticated analysis – in which
case it is possible that their definition should surround price and resource measures
rather than more cultural measures.
With reduced investment motives it may be possible to conclude that other
‘gentrification benefits’ provide as big a motive; for example, the desire for green space,
shorter journeys to work and a cheaper house centrally located. It is interesting, and
perhaps significant, that many of London’s ‘loft’ conversions are taking place in areas
not previously associated with gentrification. Ironically it is in the current property
recession that areas such as Southwark, which have always been well located vis-a-vis
access to the city of London, have only just begun to be gentrified in a property
recession. It may well be that Smith’s ideas on the ‘rent gap’ theory of gentrification are
only just being realised as these areas are clearly profitable because of the devalorisation
of land due to industrial decline.
Through an analysis of the gentrification literature one can see two continuous or
essential components to the process, these are; (i) Class or socio-economic group
movement within (ii) a discrete geographical area. This leads us to the definition:
‘Gentrification is the movement of middle and upper class households into discrete
areas occupied (or previously occupied by) by lower status (working-class)
households.’ It should also add that it is recognised that contingent features of the
phenomenon are in abundance and this is considered later on. Of course what this does
not do is explain why or when such movements should occur, however a definition
serves as a way of identification of phenomena amongst a plethora of diverse and
There is a tension between the differing scales at which gentrification may be defined.
Although there has often been a reference to the area or neighbourhood basis of
gentrification it is clear that the process is essentially constituted of households and
individuals. If one defines gentrification as an area based or group phenomenon one
loses sight of its make up and this also poses the problem that if gentrification is an area
phenomenon it makes it difficult to suggest that there is such a thing as a ‘gentrifier’
since such a role may only be accorded to someone when a neighbourhood or area
reaches a gentrified state. Similarly it is difficult to suggest a cut off point at which such
neighbourhoods shift from being non-gentrified to gentrified since this would impose a
numeric formalism that would be out of keeping with its shifting and processual
constitution even though such criteria have been set (see Phillips, 1993).
This makes gentrification a reified concept. Its structural definition belies its constitution
through human action. The only solution to this problem is to view gentrification as a
household phenomena. While it can be argued that gentrification is a neighbourhood
process it is clear that is made up of individual household movements. If one can accept
this then it is possible to acknowledge the existence of gentrifiers as the building blocks
of what can become neighbourhoods with varying degrees of gentrification rather than
gentrified areas which contain gentrifiers.
The later use of census variables relating to the occupational structure clearly clouds as
much as it elucidates such a group and is theory dependent. The groups identified stem
from both the literature and the logical selection of those occupational groups which are
higher than others. This necessarily suggests that all professionals and managers are
gentrifiers in the context of the operational research. The use of owner occupation as a
measure was fatally flawed because of the extent of right to buy policy and degree
holders were not a separable group from those with any form of higher education.
The ‘cultural’ element alluded to in the literature is difficult to operationalise and
thereby quantify. As mentioned before, it may be possible to provide some form of
measurement scale by reference to various cultural ‘artefacts’ such as antiques, coloured
doors and so on but this relates to the idea that these are strong signifiers of gentrifiers,
liable to be an erroneous assumption. It is persistently likely that theoretical ideas about
the nature of gentrification will be more complex and comprehensive than operational
measures designed for the measurement of such phenomena.
Many theorists have argued (Bulmer, 1984, Craib, 1984, Ackroyd and Hughes, 1993,
Gilbert, 1993, Ritzer, 1996) that there is a very strong link between the way one
conceive of society and social artefacts and the way research is subsequently engaged.
While not all theory is applied all research is, necessarily, theory based; it is widely
recognised that research into social life is ‘theory dependent’ and that our ‘ability to
make connections between action, experience and change is based on the explicit use of
To view research as a foundationalist enterprise in which the collection of data leads to
inevitable conclusions has come to be viewed as an erroneous picture of the progress of
research (Hammersley and Gomm, 1997). Cicourel (1964) has shown that research is
constantly mediated by theoretical concerns and social pressures, and, more importantly,
that categories are social constructs and not objects objectively available for scrutiny.
The later use of multiple regression models and statistical significance tests show that
theory is still an inherent part of what appear to be value-neutral mathematical
processes. However, it is clear that both the interpretation of results and the use of
theory in guiding the hypotheses and assumptions of such models are an essential
component (Abelson, 1995).
The chronology of gentrification and displacement
In relation to the later use of theoretical models and the better conceptual understanding
of gentrification and displacement a clear problem existed surrounding the precise
timing of both gentrification, but also displacement. Gentrification cannot, itself,
displace people, property must be vacant before moves are made by gentrifiers; in other
words, how can gentrification be seen to cause displacement. In the chronology of
events displacement must logically occur prior to any act of gentrification; it is
necessary for displacement to occur before gentrification so that dwelling spaces can be
made available. Of course it is possible that gentrifiers might purchase property with
sitting tenants and wait for them to leave but this is relatively rare.
It has frequently been held that gentrification can cause displacement but in actual fact
this can only be true in terms of an anticipation process in which developers and
landlords see that profits can be made by a change in use, occupation or through sale. It
may be that this comes from a relatively low level of activity in which some junior
professionals, students or artists promote the viability of the gentrification of an area.
This implies that such a chronology helps hide the causal relationships between the
events since displacement is absent by the time gentrification takes place, thus softening
the impression that it gives of displacing people.
Gentrification represents a self-fulfilling prophecy for the speculator landlord or
developer who both create and react to market potential and opportunity. The gentrifier
or professional’s conscience is clear, “we didn’t do it, it had already happened”, why it
happened is another question. It will later be seen that this has ramifications for the
direction of causality assumed within regression models and correlations.
It is the anticipation by developers and landlords that a profit can be made that may lead
to harassment, eviction or notices to quit. Where displacement occurs because an area
becomes too expensive the move may appear to be made on a voluntary basis,
obfuscating the real reasons for the move. Finally, it is possible that moves are also
made because of an inability to enter the market in a gentrified area. These moves may
be near impossible to empirically chart but will still be due to the same factors.
With regard to the role of preconceptions and hypotheses in the research these were
neither used to confirm or refute the existence or nature of displacement in relation to
the gentrification process in London. This cautious yet open approach reaped rewards
through its lack of orthodox assumptions. For example, in exploring the constitution
of gentrification within the capital no assumptions were made about the geographical
location of the phenomenon and this lack of guiding hypotheses (itself guided by
Spittles, 1996, “Boom market has investors sitting pretty”, The Guardian, May 26
more open hypotheses as to the nature of gentrification) yielded results which one will
later see contradicted previous research which had asserted the primacy of the
phenomenon in the inner city.