Urban Studies, Vol. 40, No. 12, 2401-2426, November 2003 jjÄ Carfax Publishing ]H9 Taylor** Francis Group Gentriflcation and the Middle-class Remaking of Inner London, 1961-2001 Chris Hamnett [Paper first received, December 2002; in final form, May 2003] Summary. This paper reviews the debates over the explanation of gentriflcation and argues that gentriflcation is best explained as the social and spatial manifestation of the transition from an industrial to a post-industrial economy based on financial, business and creative services, with associated changes in the nature and location of work, in the occupational class structure, earnings and incomes and the structure of the housing market. The paper sets out the links between these changes in the London context. It also examines the evidence for gentriflcation-in-duced displacement in London, arguing that it may be more appropriate to view the process partly as one of replacement. Introduction The term gentriflcation was first coined in 1964 by Ruth Glass to describe the changes she observed in the social structure and housing market of parts of inner London. One by one, many of the working class quarters of London have been invaded by the middle classes—upper and lower ... Once this process of 'gentriflcation' starts in a district it goes on rapidly until all or most of the original working class occupiers are displaced and the whole social character of the district is changed (Glass, 1964, p. xviii). Her use of the term 'gentriflcation', which sometimes vexes or puzzles academics who prefer the term neighbourhood 'revitalisa-tion' or 'renovation', was deliberately ironic and tongue-in-cheek. Rooted in the intricacies of traditional English rural class struc- tures, the term was designed to point to the emergence of a new 'urban gentry', paralleling the 18th- and 19th-century rural gentry familiar to readers of Jane Austen, who comprised the class strata below the landed aristocracy, but above yeoman farmers and peasants. She identified gentriflcation as a complex process involving physical improvement of the housing stock, housing tenure changes from renting to owning, price rises and the displacement or replacement of the working-class population by the new middle class. Ten years later Glass commented that London is now being 'renewed' at a rapid pace—but not on the model about which we are so often warned. Inner London is not being 'Americanised': it is not on the way to becoming mainly a working class city, a 'polarised' city, or a vast ghetto for Chris Hamnett is in the Department of Geography, King's College, The Strand, London, chris. hamnett @ kcl.ac.uk. WC2R 2LS, UK. E-mail: 0042-0980 Print/1360-063X On-line/03/122401-26 © 2003 The Editors of Urban Studies DOI: 10.1080/0042098032000136138 2402 CHRIS HAMNETT a black proletariat. The real risk for Inner London is that it might well be gentrined with a vengence, and be almost exclusively reserved for selected higher class strata (Glass, 1973, p. 423). The social and housing market changes associated with gentrification which Glass identified in inner London in the 1960s and early 1970s have since grown considerably and the importance of gentrification in dramatically reshaping the social geography of inner London over the past 30-40 years cannot be overstated. As Butler notes London is being 'made over' by an urban centred middle class. In the post war era, upwardly mobile social classes tended to leave the city. Now, led by a new middle class, they are reconstructing much of inner London as a place both in which to work and live (Butler, 1999, p. 77). The various interpretations and explanations for gentrification are well known, but the argument which will be put forward here is that gentrification is the social and spatial manifestation of the transition from an industrial to a post-industrial urban economy based on financial, business and creative services, with associated changes in the nature and location of work, in occupational class structure, earnings and incomes, life styles and the structure of the housing market. The analysis accords quite closely with that of Ley's (1980, 1981) work. It is argued that one of the key characteristics of gentrification is that it is not evenly distributed across a variety of different towns and cities, but is particularly concentrated in a relatively small number of major cities such as London, New York, Paris, Sydney, Toronto, Boston and San Francisco where the transition from industrial to post-industrial economy has been marked, where the professional and managerial middle classes have expanded and where there is an attractive old 19th- or early 20th-century inner-city housing stock suitable for renovation and conversion. Although gentrification has been found in older industrial cities such as Philadelphia (Smith, 1979) and Baltimore (Harvey, 1974), Glasgow (Bailey and Robertson, 1997), Manchester and Leeds (Dutton, 2003) during the 1990s, it is often a more recent phenomenon and is generally less marked as the changes in the industrial and occupational class structure have been slower. The Explanation of Gentrification The academic literature which tries to document and explain the rise of gentrification is very extensive (see Hamnett, 1984, 1991; Smith, 1979, 1996; Butler, 1997 and Ley, 1996, for summaries) but it hinges around three key competing explanations. The first, put forward by Ley (1981, 1996) and others, argues that the roots of gentrification lie in the changing industrial structure of major cities with the switch from manufacturing industry to service-based industries and a concomitant change in the occupational class structure from one based around the dominance of a large manual working class to one increasingly dominated by white-collar professionals, managers and technical workers in the financial, cultural and service industries which are concentrated in major cities. Secondly, and related to this, it is also argued by Ley (1980) and Butler (1997) and others that, as a result of these changes in class composition, there have also been changes in cultural orientation and preferences and working patterns of a fraction of this new middle class which have predisposed them to living in the inner city, rather than commuting from suburbia (May, 1996). Other authors (Bondi, 1991; Warde, 1991; Butler and Hamnett, 1994; McDowell, 1997) have also pointed to the importance of changing gender relations, particularly the growing importance of dual professional households. Thirdly, and strongly opposed to the previous interpretations, Smith (1979, 1987, 1996) has argued that the driving-force behind gentrification is not the new middle class (whose existence he doubts), but the growing gap between property values and underlying land values in the inner city. This GENTRIFICATION AND INNER LONDON 2403 has, he believes, opened up a growing 'rent gap' which has been exploited by the actions of property-based capital, estate agents, developers and the like, which have redeveloped or rehabilitated and gentrified undervalued inner-city housing for profit. In Smith's phrase, gentrification represents 'a back to the city movement by capital, not people'. In his view, the focus on cultural values and residential preferences of the new middle class is a diversion from the key issues which involve the structure of the land and property market and its financing, rather than demand and preference. Smith developed a sophisticated theoretical argument regarding the development of a 'rent gap' in the inner city which emerged as a consequence of major suburbanisation, and the devalorisation and abandonment of many US inner-city areas post-war. He argued that while the potential value of inner-city land remained high, the value of the buildings had fallen dramatically. Consequently, a growing 'rent gap' emerged between the potential value of the land and its existing use value. The size of the gap grew until it was possible for developers to move back to the inner city and profitably realise the underlying value of the land through renovation or redevelopment of the buildings. Smith is clearly right regarding the importance of capital depreciation and subsequent reinvestment in helping to explain gentrification, but the Achilles heel of his argument is his unwillingness to accept the significance of increased demand for inner-city locations from the expanded middle classes. Smith accepts that Western capitalist economies have experienced a decline in the relative importance of manufacturing employment and parallel increase in the importance of professional, administrative, service and managerial occupations in the producer services sector (finance, insurance, real estate and such), non-profit services (mainly health and education), and the government sector (Smith, 1987, p. 153). but he is unwilling to accept the implication of these points—namely, that gentrification is closely linked to the existence of an expanded middle class and associated demand for inner-city space. The argument made here is that the basis of an effective explanation has to rest on the demand side as much or more than the supply side of the equation. But, in making this case, an argument is not being made for a naive version of consumer sovereignty, but for recognition of the importance of changes in the economic base and class structure of cities in the transition from industrialism to post-industrialism. This shift provides the basis for an expansion of middle-class housing demand in the inner city. Smith's objection to demand-led explanations is that they are overly individualistic, place too much stress on shifts in consumer choice and preference, and fail to provide an adequate explanation of underlying changes in the land and property markets. He also argues that they are insufficiently materialist in their theoretical approach in that they fail to deal with underlying economic changes. But demand-based arguments are not just based on consumer taste and preference. In this case, they locate the basis of gentrification demand in the shifts in industrial, occupational and earnings structures linked to the shift from industrial to post-industrial cities. The addition of arguments regarding the residential preferences of a specific fraction of a growing middle class does not undermine the underlying structural economic basis of the argument. Although a demand-based explanation must also account for the reasons why housing in the inner city had become deteriorated in the first place, it is suggested that this was a consequence of a lack of effective demand from low-income working-class renters, combined with suburbanisation and the denial of mortgage finance to inner-city areas as Harvey (1974) and Williams (1976, 1978) have shown. The growth of the middle classes in cities with a growing financial and business service sector, and increasing willingness for institutions to lend in inner-city areas, permitted gentrification to take off. 2404 CHRIS HAMNETT There is also another explanation put forward by Redfern (1997) which argues that gentrification took place because of the availability and falling real cost of domestic technologies which permitted old, 19th-century houses to be modernised and upgraded to 20th-century standards. He argues that gentrification rests on the ability to modernise houses. This is an important point but it seems to be a necessary, rather than a sufficient factor for gentrification to occur. If the supply of gentrifiable properties and the demand for them were not there, domestic technology alone would be unlikely to lead to gentrification. The Transformation of London's Economic Base The economy of London has undergone a dramatic transformation over the course of the past 30^10 years. Until the mid 1960s, it was a major centre of British light industrial production and a third of its labour force were employed in manufacturing. Although it had long functioned as an important national and international financial centre, the importance of finance and business services for overall employment was relatively small until the 1980s. In 1961, only about 1 in 10 of all London's workers were employed in this sector. But in the past 40 years, the proportionate importance of manufacturing industry and finance and business services has been totally reversed. London's economy has undergone a dramatic transformation from an industrial city to a post-industrial city dominated by financial and business services and the creative industries (Clark, 2002; Pratt, 1997; GLA Economics, 2002). Manufacturing industry and employment have been reduced to a vestigial remnant of their former importance. Because of its leading role as major international financial centre, London contains a disproportionate concentration of jobs in financial services. Secondly, because of its role as international headquarters of major corporations, it contains a high proportion of jobs in corporate headquarters and in related business services such as management consultancy, law, advertising, design and public relations. Thirdly, because of its role as a centre of broadcasting and publishing, it has a disproportionate concentration of jobs in media production and marketing, ranging from national newspapers and journals to the major TV studios, record and video production. Finally, particularly in the 1990s, it has become a major centre for the cultural and creative industries. In 1998, finance and business services together employed almost a third of London's workers, whereas manufacturing industry employed only 8 per cent, and an estimated 50 per cent of London's manufacturing employees worked in the head offices of manufacturing companies. These changes are reflected in the employment figures by industry. In 1961 London had 1.45 million manufacturing jobs (32.4 per cent of the total). By 1991 the total had fallen to 359 000 (11 per cent) and to 253 000 (8 per cent) in 2001. The number of manufacturing workers stabilised in the 1990s at around 288 000 or 8 per cent (Tables 1 and 2). The growth sectors in the 1980s and 1990s have been finance, business services and the creative industries. Business services are now an extremely important element in the economy and labour force of global cities and London is no exception to this trend. Over the period 1981-98, employment in business services in London increased by 376 000 or 92 per cent, from 408 000 (11.6 per cent of total employment) to 784 000 (22.1 per cent). From just 1 in 10 of jobs in Greater London in 1981, the business and professional services sector accounted for 1 in 5 jobs by 1998 (Daniels, 2000). By 1999, finance and business services accounted for 32 per cent of total employment, more than any other sector and four times that of manufacturing industry. This is a total reversal of the position in 1961 when manufacturing employed more than three times the number of workers in finance and business services (see Tables 1 and 2 and Figure 1). A report by GLA Economics (2002) suggests that creative industries (including fashion, advertising, digital arts and advertising) are now the Table 1. The employment structure of Greater London by industry, 1981-91 Employment chang ;e by sector Chang e in employment by 1961 1981 1991 industrial sector, 1981-91 Percentage Percentage S Number Percentag e Number Percentage Number Percentage Number change point change H Primary industries 93 400 2.1 57 275 1.6 41364 1.3 -15911 -27.8 -0.3 Manufacturing industry 1 453 000 32.7 683 951 19.2 358 848 11.0 - 325 103 -47.5 -8.2 p Construction 280 900 6.3 161407 4.5 118 367 3.6 - 43 040 26.7 -0.9 H O 2 Distribution, hotels and 734 700 16.5 686 598 19.3 645 955 19.8 - 40 643 -5.9 0.5 catering, repairs > 2 Transport and 433 700 9.8 368 288 10.3 307 682 9.5 - 60 606 -16.5 -0.8 O communications 2 Banking, finance, 461 800 10.4 565 876 15.9 733 513 22.5 167 637 29.6 6.6 M insurance, business r services o o o Other services 989 800 22.2 1034 526 29.1 1049 015 32.2 14 489 1.4 3.1 Total services 2 620 000 58.9 2 655 288 74.6 2 736 165 84.1 80 877 3 9.5 2 Total employment 4 447 300 100 3 560 688 100 3 254 744 100 - 305 944 -8.6 0 Sources: Census of Population, 1981 and 1991. 2406 CHRIS HAMNETT Table 2. Employment change in London during the 1980s and 1990s Job change 1993-98 1983-88 Finance 25 000 58 000 Business services 210 000 126 000 Hotels and restaurants 56 000 16 000 Manufacturing 0 - 142 000 All other 179 000 34 000 Total job change 470 000 92 000 Percentage change 15.2 2.7 Source: Banks and Scanlon (2000) from Annual Employment Survey. major employer in London after business services, although they are not identified in the standard industrial groupings. The Transformation of London's Occupational Class Structure The transformation of London's industrial structure has been paralleled by changes in its occupational class structure. Contrary to the proletarianisation and polarisation theorists, the census and General Household Survey show that London has witnessed a substantial growth in the number and proportion of its managerial and professional workers. The evidence for this has been discussed in detail elsewhere (Hamnett, 1994, 1996) and it is unnecessary to duplicate this here. Suffice to say that the census data for both economically active males (employed and unemployed) from 1961 to 1991 and economically active females from 1971 to 1991 show a significant and consistent growth in the proportion of professional and managerial groups and a significant and consistent decline in the size and proportion of skilled, semi-skilled and unskilled manual workers. There has been a fundamental change in the occupational class structure of London over a 30-year period which is consistent with the shift from a manufacturing to a financial and business service-based economy. The traditional and long-standing man- ufacturing manual working class has shrunk and has been replaced with a large service-based middle class (see Figures 2 and 3). Critics such as Brueghel (1996) and Cox and Watt (2002) argue that the picture presented by census data is flawed and partial in that it overlooks the substantial number of workers in part-time, illegal or undocumented employment in catering, cleaning, hotels, personal services and the like. There is validity in this argument, but there is no systematic evidence to support it. It should also be pointed out that the argument for the existence of a large undocumented personal service sector tends to overlook the fact that there have been large scale losses of low-skilled service jobs in transport and other sectors. The large numbers of railway porters, rail and tube guards and bus conductors of 30 years ago have all but disappeared, although this tends to be ignored by the proponents of a low-skill service economy thesis (Tables 3 and 4). There are also low-skilled jobs in financial and business services, but the overall trend is one of professionalisation. The Transformation in Earnings and Income Structure The transformation in the industrial and occupational class structure of London has been paralleled by marked changes in the GENTRIFICATION AND INNER LONDON 2407 ■ 1961 % ■ 1981 % D 1991 % G 1998% 1 r^-u. , business Transport and Health. Education Manufacturing Other services industry Primán/ and Figure 1. The changing employment structure of Greater London, 1961-98. structure of both individual earnings and household incomes. Looking first at individual earnings, data from the New Earnings Survey 1979-95 show that there has been a marked upwards shift in the distribution, although the survey is based only on those paying tax and therefore underrepresents low-income part-time workers. Table 5 shows that mean gross weekly pay for full-time males in Greater London rose by 57 per cent in real terms compared with 47 per cent in the South East and 35 per cent in Great Britain as a whole (GB). For full-time females, earnings rose by 73 per cent in Greater London, 65 per cent in the South East and 56 per cent in GB. Earnings of female part-time workers rose by 35 per cent in GB, 41 per cent in the South East and 45 per cent in Greater London. Whereas male mean weekly pay in Greater London was 14 per cent greater than for GB overall in 1979, by 1995 it was 33 per cent higher. The equivalent figures for full-time females were 16 per cent and 29 per cent respectively. The differences in median earnings were lower but show the same pattern. Earnings in Greater London increased at a much faster rate than in both Britain and the rest of the South East. This reflects London's role as a global city and its industrial and occupational composition. In addition, the proportion of high earners as a proportion of all earners also rose sharply from 1979 to 1995. The rich are not only getting richer, but there are more of them and the gap between the earnings of the higher groups and the rest has grown, leading to an increase in inequality (Hamnett and Cross, 1998a) Table 5 shows that earnings in both the top decile and the top quartile rose far more rapidly than in the bottom decile and quartile. Earnings of the top decile of full-time males in London rose by 73 per cent over the period while the earnings of the bottom decile rose by 18 per cent: a ratio of 4:1. For female full-time employees, earnings in the top decile increased by a remarkable 83 per cent but even in the bottom decile it was still 45 per cent: higher than for males. Earnings of the top decile of female part-time workers in London rose by no less than 97 per cent, although these earnings are much lower than the other two groups. This argument that the growth of inequality is driven by earnings growth at the top is strongly supported by Gordon (2000) who notes that Earnings inequalities in the London labour 2408 CHRIS HAMNETT Intermediate and Skilled manual Semi-ski led Unskilled Armed lorces and junior non- manual occupation manual inadequately described Managerial Figure 2. The changing distribution of socioeconomic groups in Greater London, 1961-91: economically active males. market are greater than in the country as a whole, although this is almost entirely due to the top tail of the earnings distribution among male non-manual workers (where the top decile was 75 per cent above the average, and 4.3 times the lowest decile, compared with 3.7 times nationally). In other words, it reflects the presence of a disproportionately large number of very highly rewarded professional and managerial workers, rather than any larger number of poorly paid jobs (Gordon, 2000, pp. 19-20). Sassen (1991), Friedmann and Wolff (1982) and Mollenkopf and Castells (1991) argue that the financial and advanced business service sectors are one of the major motors of increases in earnings inequality in global cities. Not surprisingly the NES data show that earnings are higher in the City of London and Westminster (the two main centres of financial and business services employment in London) and have risen faster than elsewhere. Table 6 shows that mean earnings rose by substantially more for both males (100 per cent) and females (102 per cent) in the City of London than they did in London as a whole (57 per cent and 73 per cent respectively). This was also true for the top and bottom deciles of earnings. Earnings for the top decile of male full-time workers rose by 136 per cent in the City compared with 73 per cent in London and female full-time earnings rose by 117 per cent compared with 83 per cent. The City has led the upwards growth in earnings in London. To this extent, it is also a major driver in the growth of earnings inequality which increased substantially in London over the period. The latest NES data for 2001 confirm the continuation of this picture of rapidly rising earnings at the top end and growing inequality. Household Income Change It is also important to examine the changes in the level and distribution of household income, not least because the household is the key economic unit for housing costs. Between 1979/80 and 1989/90 average household incomes increased in real terms by 37 per cent in London compared with an average of 23 per cent in the UK. Incomes in London are rising more rapidly than in the rest of the country and have been since the GENTRIFICATION AND INNER LONDON 70 60 - 50 - g, 40 to I 30 20 - 10 D 1971 (%) D 1981 (%) n 1991 (%) m 1,2,3,4 5,6 8,9,12,14 7,10,15 11, 16,17 Figure 3. The changing distribution of socioeconomic groups in Greater London, 1971-91: females. late 1970s. The shape of income distribution between London and the South East and the rest of Britain is also different. Comparing income distributions for London and the UK for 1981 and 1995-98 shows that London had 50 per cent more households (21 per cent against 14 per cent) in the highest income category than the UK. London had 31 per cent of households in the top two groups in 1996-99 compared with 23 per cent in the UK as a whole (Table 7). This has an inevitable impact on the structure of housing demand. There is also a growing disparity between no-earner households on the one hand, and dual-earner households, particularly dual professional households, on the other (Pahl, 1988). Comparing GHS data for 1979 to 1993, the median incomes of households in London with no earners rose 25 per cent in real terms, while incomes of households with one earner rose by 38 per cent and those with two earners rose by a remarkable 60 per cent. Two-earner households in London also saw the greatest increase in real incomes (77 per cent). Dual-earner professional and managerial households (DPMs) comprise the top level of the household income distribution. In London, the proportion of such households trebled from 1.5 to 4.8 per cent and they saw large increases in real incomes over the period. By 1993, median DPM household incomes were £1000 per week, well above the £657 of all two-earner households and four times the median for all households in London. The growth of high incomes and growing inequality have an important effect on the housing market as is argued below. The Impact of the Growth of High Earnings and Incomes on the Housing Market In market economies, private housing is a commodity produced, sold and exchanged for profit. Outside the social or public housing sector where housing may be allocated on the basis of some criterion of need or at a below-market price, housing is competed for through the market, where it is allocated on the basis of price and ability to pay. For most buyers (60 per cent of owners in London own with a mortgage), access to private housing is significantly constrained by earned income and hence by the position of 2410 CHRIS HAMNETT Table 3. Socioeconomic change in London, 1961-91: economically active males Percentage Percentage Percentage point change point change 1961 1971 1981 1991 1961-91 1961-91 Managers 11.9 13.9 16.1 20.3 8.4 10.3 Professionals 4.8 6.0 6.4 8.2 3.4 9.4 Other non-manual 23.0 23.5 22.0 23.7 0.7 -33.2 Skilled manual 34.8 32.8 30.2 27.6 -7.2 -48.7 Semi-skilled manual 14.5 13.4 14.3 12.4 -2.1 -44.8 Unskilled manual 8.3 7.2 5.8 4.4 -3.9 -65.3 Armed forces/other 2.6 3.1 5.2 3.2 0.6 -20.5 Total 100 100 100 100 0 0 individuals within the paid labour market (Hamnett and Randolph, 1988). Given this relationship, the argument here is straightforward. First, that position in the London housing market is strongly related to labour market position and income, and, secondly, that the changes in London's industrial, occupational and earnings structure have had a dramatic impact on the nature of the housing market. Put simply, the sharp increases in the size of the professional and managerial occupational groups over the past 30 years—combined with the very large increases in earnings and incomes that this group have had—have, along with the declining numerical importance of manual workers and the growth of economic inactivity, significantly changed the structure of housing market demand in London. This is not to suggest that all, or even a majority, of professional and managerial high-income households in London will be gentrifiers, but that the expansion of this group will increase the pressure of housing demand. Merrett (1988) argues that owner-occupied housing in London can be characterised in Ricardo's terms as one of those commodities 'the value of which is determined by their scarcity alone'. Location is also a key element of scarcity. Such commodities have the attribute that the supply can only be increased slowly and with considerable difficulty and cost. In these circumstances, price is largely determined by demand. Increase demand (or the ability to pay) and the price will rise accordingly. This has been intensified by London's role as a centre of global finance and the growing number of small households. Not surprisingly, mean house prices have risen dramatically over the past 30 years and more rapidly than in Britain as a whole. In 1970, average prices in London were 120 per cent of the UK average. By 2001, they were 165 per cent of the UK average. Although prices in London fell sharply in the housing market slump of the first half of the 1990s (Hamnett, 1993, 1999), they have recovered strongly and average prices in London rose by 137 per cent from June 1995 to June 2002: from an average of £98 300 to £232 800. Significantly, percentage increases in recent years have been much higher in central and inner London boroughs than in outer London. At the top end, 3 of the top 5 boroughs with increases of 178 per cent or more were in the East End (Newham, Tower Hamlets and Hackney) and almost all boroughs with increases of 140 per cent or over were in inner London (Table 8 and Figure 4). This is indicative of the extent of gen-trification in inner London. The growth of house price inflation has created major problems for less affluent households, however. With average incomes in London of £22 000 per year, it is virtually impossible for households with incomes under £30 000 a year to buy even in the cheapest boroughs. As shown previously, London has experienced a dramatic increase in both the size GENTRIFICATION AND INNER LONDON 2411 (N O O (N I in a\ oo 3 o u O -O C O *0 C O J oo c Oh 3 2412 CHRIS HAMNETT Table 4. Socioeconomic change in London, 1971-91: economically active females Percentage 1971 1981 1991 1971-91 point change, 1971-91 Managers/professionals Other non-manual 6.3 56.4 9.3 58.5 16.9 58.3 10.6 1.9 140.4 -7.4 Skilled manual 6.3 4.5 4.4 -1.8 -36.9 Semi-skilled manual 20.4 18.0 13.1 -7.3 -42.5 Unskilled manual 6.2 5.1 4.9 -1.3 -28.7 Armed forces/other 4.4 4.6 2.4 -2.0 -51.6 Total 100 100 100 0 -10.4 Hamnett (1996) from Censuses of Population, 10 per cent tables. and proportion of its professional and managerial labour force since 1961 as a result of the transformation of its economic base from manufacturing industry to services. This expanded middle class needed somewhere to live. The private rented sector was beginning to contract rapidly as a result of local authority redevelopment and by sales of better-quality property to owners. The choices were thus threefold: to buy outside London and commute in to work, to buy in the outer London suburbs and commute in to work, or to rent/buy in inner London preferably in an area close to the employment opportunities and entertainment facilities of central London. A significant proportion opted to live in inner London (McDowell, 1997), but this growth in demand could not be accommodated in middle-class residential areas where prices began to rise rapidly. The solution, perceived by early pioneers in the late 1960s, was to turn to the areas of well-built, aesthetically attractive, centrally located but multiply occupied, decaying, but low-priced areas which had been abandoned by the middle classes and taken over by working-class populations in previous decades. These houses, often large and well laid out, with high ceilings, large rooms and attractive architectural features, offered a large amount of space per pound and the middle classes were very willing to acquire cheap and potentially desirable homes. Michael Frayn (1967) perceptively captured this in his novel Towards the End of The Morning They decided to find a cheap Georgian or Regency house in some down at heel district near the core. However, depressed the district, if it was Georgian or Regency, and reasonably central, it would soon be colonized by the middle classes. In this way they would secure an attractive and potentially fashionable house in the heart of London, at a price they could afford; be given credit by their friends for going to live among the working classes; acquire very shortly congenial middle-class neighbours of a similarly adventurous and intellectual outlook to themselves; and see their investment undergo a satisfactory and reassuring rise in the process (Frayn, 1967, p. 73). On the supply side, private landlords who had been squeezed by years of rent control and low prices, were only too willing to take advantage of the new demand and higher prices. As Williams (1976, 1978) has shown, building societies who were previously very unwilling to lend on inner-city housing, also began to relax their lending criteria. As a result, potential buyers found it far easier to gain access to mortgage finance and the new middle class began to expand residentially into areas which were previously predominantly working-class. The supply of cheap period houses for conversion to single-family residences in the GENTRIFICATION AND INNER LONDON 2413 Table 5. Percentage increase in gross weekly pay, 1979-95 (at April 1995 prices) for Great Britain/GB, the South East (SE) and Greater London (GL) Male full-time Female full-time Female part-time workers workers workers GB SE GL GB SE GL GB SE GL Top decile 48.6 62.4 73.1 69.5 76.5 83.4 82.3 97.3 97.2 Top quartile 37.3 48.3 55.9 66.4 72.9 77.9 45.9 55.8 65.5 Mean 34.6 46.9 57.2 56.1 65.5 73.5 44.8 55.7 60.9 Median 25.4 34.4 41.1 48.0 58.7 46.5 28.9 35.3 44.1 Bottom quartile 14.6 22.2 27.0 36.6 46.5 54.7 16.2 22.9 27.8 Bottom decile 7.1 11.7 17.8 26.6 36.1 45.4 10.8 14.3 16.7 Sources: New Earnings Survey, 1979 and 1995. more attractive parts of the inner city dried up rapidly, aided by the four house price booms of the early and late 1970s, the second half of the 1980s and the second half of the 1990s. As a consequence, potential gen-trifiers had to look further afield within London or resort to smaller properties or converted flats. The conversion of older houses into flats emerged in the 1980s as developers saw the profits to be made. By the end of the 1980s, conversions were the single largest source of new dwellings in London (Hamnett, 1989). More recently, the conversion of old warehouses, factories and offices has pushed the process into city-fringe areas such as Hoxton, Clerkenwell, Shoreditch and Whitechapel (Hamnett and Whitelegg, 2001) while Docklands has undergone comprehensive regeneration. In the process, the social geography of inner London has been dramatically changed. Gentriflcation and Social Change in Inner London The importance of gentriflcation in dramatically reshaping the social geography of inner London over the past 30^10 years cannot be overstated. Much of inner London in the 1950s and 1960s was dominated by a large working-class population. The social geography of London is shown by a cluster analysis of 1966 census data undertaken by the Greater London Council (GLC) (Daly, 1971) using 11 selected census variables on occu- pation, tenure, household structure and mobility for each of Greater London's 7000 wards. Figure 5 shows a sharp contrast between the East End and the West End of London, with the former characterised by high levels of council housing and working-class residents, and the latter characterised by high levels of young, mobile, middle-class private renters. Surrounding central London, much of inner London was characterised by working-class private rented housing surrounded by the more middle-class owner-occupied suburbs. What gentriflcation has done is to transform much of inner London in both social and tenure terms. In the process, the remaining working-class residents of inner London have been squeezed into residual areas. Smith and Williams (1986) have termed this the victory of the 'space invaders'. In 1961, professionals and managers were concentrated in a few areas of the central boroughs of Camden, Westminster and Kensington, but in each successive decade the distribution spread out into surrounding inner-city areas. This is very clearly seen in Figure 6 which shows the percentage change in the proportion of professionals and managers by borough, 1981-91. The boroughs with the highest rates of increase were Tower Hamlets (72 per cent), Wandsworth (63 per cent), Hammersmith and Fulham (57 per cent) and Islington (56 per cent), with other large increases in Southwark (46 per cent), Lambeth (43 per cent) and Hackney (41 per , Greater London: Location of wards in each cluster at 1:63360 scale ^T'^wpT^ijLj.i'ffiuwy'.^t' '.vi"-^-*'m^^rr' •z n Figure 5. Social variations in Greater London (based on cluster analysis of 1966 census data). H H O > O 2 M r o o o •z Figure 6. Percentage change in number of professional and managerial residents in London boroughs, 1981-91. 2416 CHRIS HAMNETT Table 6. Percentage increases in gross weekly pay and interdecile ratios, 1979-95 (at April 1995 prices) for Greater London (GL), Westminster (West) and the City of London(City) Male full-time workers Female full-time workers GL West City GL West City Percentage increases in gross weekly pay Top decile 73.1 68.1 135.8 Mean 57.2 65.6 99.9 Bottom decile 17.8 23.7 43.1 Interdecile ratios 1979 2.6 3.17 2.98 1995 3.82 4.3 4.9 Change 1.22 1.13 1.92 83.4 73.5 45.4 2.26 2.85 0.59 93.7 73.6 48.5 2.13 2.78 0.65 116.6 102.3 74.0 2.11 2.62 0.51 cent). In some boroughs, such as Tower Hamlets, the absolute increase is less marked, but it is still significant. When the percentage change in the number of professionals and managers by borough 1981-91 is plotted against the proportion of professionals and managers by borough in London in 1981 (Figure 7), the greatest percentage increases in the professional and managerial group tended to occur in those boroughs with the lowest percentages in 1981. Although the pattern is not perfect, correlation analysis gives a multiple R of 0.59 and an R2 of 0.34. This indicates large-scale gentrification of previously working-class areas during the 1980s. A clear divide exists between the Inner London boroughs with a mean increase of 38 per cent in the size of the professional and managerial group and outer London boroughs with an increase of 21 per cent. The East End is still predominantly working-class but the changes are indisputable. Arguably, the reason the central boroughs of Westminster, Camden and Kensington do not show an increase in the proportion of professionals and managers is that they were already substantially gentrified and very expensive. The sequence of middle-class colonisation can be traced outwards: north across most of Camden, Islington and western Haringey; east into Hackney, Stoke Newington and Docklands; west into Hammersmith and Fulham, North Kensington and Ealing; north-west in Kilburn; south into Vauxhall, Bermondsey, Clapham and the whole river strip, and finally back into inner areas of the East End adjacent to the City of London (Hamnett and Williams, 1980; Munt, 1985; Bridge, 1994; Butler, 1997; Glancy, 1999). To understand the spatial expansion of gentrification and its links to the rise in house prices, it is useful to take the analogy of a multibowl water fountain. The volume of water at the top represents middle-class housing demand. The water falls into the top bowl but, as prices rise, this is soon filled and the water spills over into the next bowl which in turn spills over down to the lowest and broadest bowl of the fountain. In this way, gentrification has spilt over into successively wider geographical bowls as prices have risen and it has an influence outside London (Thrift and Ley shon, 1992). In this respect, gentrification can be seen as a form of spatially diverted middle-class housing demand. Ley (1981) has suggested that gentrification may lead to a shift of the socio-spatial structure of the late 20th-century city back in time, from the industrial pattern of the 19th and first 70 years of the 20th centuries which was dominated by an expanded working class, to a post-industrial pattern which is, in many ways, akin to the pre-industrial pattern with the middle classes in the core. GENTRIFICATION AND INNER LONDON 2417 Table 7. Household income distribution, 1981, 1995-98 1981 1995-98 1996-99 Weekly income ------------------------ ------------------------ ----------------------- £s 1995/98 UK London UK London UK London Under £100 12 10 14 15 12 12 £100-150 11 11 11 11 10 9 £150-250 17 16 16 12 16 13 £250-350 18 16 13 13 13 12 £350^150 15 13 12 10 11 10 £450-600 14 17 13 14 14 13 £600-750 7 10 8 9 9 10 Over £750 5 8 12 17 14 21 Total 100 100 100 100 100 100 Sources: Gordon (2000) and Family Expenditure Survey, Office for National Statistics. The Social Consequences of Gentriflcation The social consequences of gentriflcation are predictable. In a competitive housing market where access is ruled by price, the expansion of the middle classes in inner London has been associated with the rolling-back of the less skilled, the unemployed, the poor and ethnic minorities who have been steadily concentrated into the remaining inner London local authority estates and the growing housing association sector. Once again, the process was perceptively anticipated by Ruth Glass who noted that: There is very little left of the poorer enclaves of Hampstead and Chelsea: in these boroughs, the upper-middle-class takeover was consolidated some time ago. The invasion has since spread to Islington, Paddington, North Kensington—even to the 'shady' parts of Notting Hill—to Bat-tersea, and to several other districts, north and south of the river. The East End has so far been exempt, although before long some of its districts, too, are likely to be affected. And this is an inevitable development, in view of the demographic, economic and political pressures to which London, and especially central London, has been subjected (Glass, 1964, pp. 138-139). Glass went on to point out that Any district in or near London, however, dingy, or unfashionable, is likely to become expensive, and London may quite soon be a city which illustrates the principle of the survival of the fittest: the financially fittest, who can still afford to work and live there (Glass, 1964, pp. 140-141). The result of these two processes may have resulted in a decline in social segregation at the borough scale in London. There is no longer a small middle-class residential area in central London surrounded by a largely homogeneous sea of working-class housing. The social class composition of inner London is now far more mixed than it was 30 years ago. But, at the local level, it is likely that segregation has risen between wealthy home-owners in one street and low-income council tenants a few streets away. As Glass pointed out 30 years ago The 'colonising' drive of higher classes in London has been accelerated; and so working-class quarters are becoming more constricted. Apart from the very rich, it is mainly the young members (or aspiring members) of the middle-upper strata, single people or couples without family responsibilities, who are prepared to pay the exorbitant housing prices of the inner sector. As these people live mainly in one- or 2418 CHRIS HAMNETT Table 8. Mean house prices (£) and absolute and percentage price changes in London, by borough, June 1995-June 2002 Percentage Absolute change April-June April-June 1995 2002 1995-2002 1995-2002 Lambeth 76 847 228 904 152 057 197.9 Kensington 218 995 629 021 410 026 187.2 Newham 48 305 136 023 87 718 181.6 Tower Hamlets 76 168 214 095 137 927 181.1 Hackney 71060 197 716 126 656 178.2 Hammersmith 133 464 355 350 221 886 166.3 Southwark 84 154 222 500 138 346 164.4 Wandsworth 110 583 290 737 180 154 162.9 Brent 77 645 203 880 126 235 162.6 Waltham Forrest 56 463 148 098 91635 162.3 Lewisham 60 012 154 975 94 963 158.2 Westminster 181 375 468 378 287 003 158.2 Haringey 85 409 220 393 134 984 158.0 Islington 118761 290 779 172 018 144.8 Greenwich 70 664 171 688 101 024 143.0 Redbridge 73 872 179 240 105 368 142.6 Merton 93 995 227 709 133 714 142.3 Camden 148 412 358 172 209 760 141.3 Kingston 94 295 226 170 131875 139.9 Harrow 88 596 209 748 121152 136.7 Hounslow 88 673 209 889 121216 136.7 Ealing 93 554 218 835 125 281 133.9 Croydon 72 090 167 146 95 056 131.9 Barnet 106 571 246 180 139 609 131.0 Sutton 78 708 175 363 96 655 122.8 Bexley 66 178 145 629 79 451 120.1 Barking 50 378 110 158 59 780 118.7 Richmond 143 128 312 926 169 798 118.6 Enfield 76 613 167 323 90 710 118.4 Hillingdon 83 008 177 672 94 664 114.0 Havering 74 233 157 864 83 631 112.7 Bromley 96 214 200 403 104 189 108.3 Greater London 98 300 232 830 134 530 136.9 Sources: Land Registry Statistics, 1995 and 2002. two-person households, they have a disproportionately large number of households, and a disproportionately large housing demand, in relation to the total population size of their group ... There are few hurdles in their path. The working-class population of Inner London ... find it more difficult to resist being displaced or hemmed in (Glass, 1973, pp. A25-A26) She went on to add that developing in inner areas which were previously almost wholly working class, and which have now acquired a quite substantial middle-upper class component. But this is likely to be quite often only the first stage of the gentrification process, whereby the working class will be reduced, before long, to the status of a statistical minority in such areas (Glass, 1973, p. 425). A new kind of apparent class mixture is Glass was remarkably prescient in her ability GENTRIFICATION AND INNER LONDON 2419 RO 70 - 60 b0 g> 40 3 c 8 £ 30 a. 20 10 0 -10 □ % Professional and Managerial, 1981 J % Change in PM, 1981-91_________ 1 LU I ě c >. c c c c c .c. o <Ľ c o c o c ■- u O) c l/l c 0) I U) -c >. c Ü £ (D "C en X % ¥ O) í c d) 1 O 3 í f o CO ! I Figure 7. Professional and managerial workers, 1981 v. percentage change in professional and managerial workers, 1981-91 (by boroughs). to predict the future of social change in inner London. As she foresaw, gentrification has been accompanied by the departure or squeezing out of the working class from many of its traditional inner-city bastions by smaller middle-class households. It is important to ask, however, to what extent the working class were actively displaced or replaced as they slowly shrunk in numbers. Gentrification and Displacement in London There is a consistent assumption in the literature that gentrification is a direct cause of working-class displacement. While this is undoubtedly true in some cases, it is argued here that the slow reduction of the working-class population in many inner-city areas is, in part, a result of a long-term reduction in the size of the working-class population of London as a whole (by a combination of retirement, death, out-migration or upward social mobility) and its replacement by a larger middle-class population. In other words, the key process may be one of replacement rather than displacement per se. A valuable attempt has recently been made by Atkinson (2000a and 2000b) to measure the extent of gentrification-induced displacement in London using 1981 and 1991 ward-level census data. He used ward changes in the proportion of professionals and managers as a proxy measure for gentrification and a variety of proxy measures for displacement, including a decline in size of the working class, the unskilled, households privately renting, ethnic minorities, the unemployed, the elderly and lone parents. His results show that the increases in the professional and managerial group were associated with declines in all the proxy displacement variables 2420 CHRIS HAMNETT Table 9. Mean percentage point changes for gentrification and displacement variables by ward quartiles, 1981-91 Variable Top quartile Second quartile Third quartile Lowest quartile Professionals 15.1 11.06 8.93 7.26 Working class - 17.69 - 15.84 -14.7 - 13.23 Renting -6.12 -6.12 -3.75 -3.24 Elderly -3.04 -2.28 -1.71 -1.75 Unskilled -3.01 -1.82 -1.23 -1.19 Ethnic minority -0.66 -0.06 0.54 1.19 Lone parent 3.91 2.94 3.96 3.54 Unemployed 12.59 10.88 11.82 11.58 Source: Atkinson (2000b, Table 1). with the exception of unemployment and lone parents, both of which show increases across the board as would be expected (Table 9). Atkinson (2000b) outlined a regression model in an attempt to measure the statistical link between the proxy variables for gentrification and displacement, although, as he points out, the model is inverted in that it uses gentrification as the dependent variable and the 'displacement' variables as independent variables. He argues that While gentrification was the motor of displacement, logic dictates that displacement would have to take place first in order to vacate dwellings for gentriflers to subsequently occupy (Atkinson, 2000b, p. 292). The model had a high degree of statistical 'fit': an R2 value of 0.74 and a multiple correlation coefficient of 0.86. He found that decline in the working class was most important (- 0.70), followed by a decline in the unskilled (- 0.34) and unemployment (-0.17) and concluded there is: "a high degree of correspondence between the events labelled as gentrification and displacement" (p. 293). There is, however, a fundamental problem with Atkinson's analysis. Although he is correct that the increase in numbers of professional and managers in many wards has been associated with a decline in the working-class and unskilled population, private renting and the elderly, and although the increase in professionals and managers in the top quartile of wards is associated with a greater proportionate decline in the working class and unskilled and private renting, it is difficult to claim that these changes are necessarily associated with displacement in a direct causal way. As he himself notes It was hypothesized that if a negative correlation existed between these changes (ie gentrification goes up and the incidence of those termed displacees goes down) some link existed between the two events though such connection cannot be conceptualized as causal or direct (Atkinson, 2000b, p. 292; emphasis added). In a more detailed, longitudinal analysis of gentrification, Atkinson (2000a) examined the flows into and out of gentrifled areas, by grouping the gentrifled wards into a number of larger, aggregate areas. He found, analysing the structure of net flows in terms of gains and losses, that gentrifled areas were characterised by a large gain in the number and percentage of professionals, and large net losses of the inactive, the working class and the elderly (Table 10). He argues that, while we cannot know whether the net flows were the result of rental increases, landlord harassment or the decision to move on it would seem self-evident that such flows cannot be wholly due to choice or social change. It is likely that displacement effect has GENTRIFICATION AND INNER LONDON 2421 Table 10. Net flows and percentage gains or losses for all' G' areas between 1981 and 1991 Variable Net change for all 'G' areas Percentage gain/loss Professionals 18 800 20 Inactive - 38 500 - 46 Working class - 19 300 - 38 Elderly - 23 200 - 38 Source: Atkinson (2000a, Table 12). been clustered in areas where professional-isation has occurred (Atkinson, 2000a, p. 163). This is undoubtedly true, but the links do not necessarily point to large-scale displacement. On the contrary, it is suggested that, where changes in class composition are concerned, what Atkinson's data reflect is the double-edged nature of a process of city-wide industrial, employment and class restructuring. This involves a gradual contraction of the working class and its replacement by an expanded middle class rather than displacement per se (Savitch, 1988; Ley, 1981). While there is undoubtedly gentrification-induced displacement of the working classes (direct and indirect), it is important not to ignore the long-term changes in industrial and occupational structure in London over the past 40 years. In particular, the long-term decline of manufacturing employment has been a cause of the decline in number of skilled and semi-skilled manual workers, just as the growth of financial, legal and business services has been associated with the growth of the professional and managerial group. If it is assumed that the average working life of an individual is 40 years (from age 20 to age 60), the entire labour force will turn over and be replaced during a 40-year period, and a quarter will be replaced every 10 years. Consequently, a high proportion of the manufacturing labour force in 1961 will have disappeared by 1991 or 2001. They will have either retired, moved out or died. This change will have taken place as a result of long-term industrial and occupational change, not necessarily as a result of gen-trification per se. It is therefore not very surprising that there has been a marked decline in the working-class population in most wards in London. While this is likely to have been higher in wards which have witnessed rapid increases in the proportion of managers and professionals (as Atkinson's figures show), by no means all the decline in the size of the working-class population can be linked to gentrification. Indeed, even in the lowest quartile of wards ranked in terms of professionalisation, there was a large percentage point reduction in the size of the working-class population not much less than the decline in the top quartile of wards and far greater than the percentage point increase in professionals and managers (see Table 9). This suggests that the decline of working-class residents has been common across the whole of London and is not specific to gen-trifled areas, although it is more strongly marked in those areas. There has also been a major long-term decline in the proportion of private renters in London. The decline in the sector began in the 1950s and has taken place right across London, both inner and outer. In inner London, the proportion of private renters fell from 64 per cent of in 1961 to 28 per cent in 1991, and in outer London from 30 per cent to 14 per cent. The decline in private renting took place across all boroughs and it was particularly marked in Inner London in the 1960s and 1970s where large areas of private rented housing were cleared for local authority redevelopment in boroughs such as Islington and Lambeth and as a result of a shift in the economics of the housing market from private renting towards ownership (Hamnett and Randolph, 1988). It is not found in gen- 2422 CHRIS HAMNETT trifled areas alone, although the decline is likely to have been greater there partly as a result of a higher proportion of private renting initially. It should also be noted that not all working-class residents who have left gentrifled areas have necessarily been squeezed out. While many will have, some working-class owners, including ethnic minorities, may have taken the opportunity of rapidly rising prices to sell up and move out. In conclusion, while gentriflcation induced displacement undoubtedly plays a role, the decline of the working-class or private rented housing cannot be attributed primarily to displacement. It is also necessary to look at replacement as a result of a long-term process of occupational and housing tenure change. The Future of Gentriflcation Critics of the importance of gentriflcation in North America (Berry, 1985; Bourne, 1993) have dismissed it as of limited importance compared with suburbanisation and inner urban decline and argue that it is small scale and a temporary product of the post-war baby-boom generation. Others such as Bad-cock (1995) and Wyly and Hammel (1999) see it rather as a continuing phenomenon which is reshaping the inner areas of many large cities. It is argued that gentriflcation will continue to grow in importance in key cities as long as the industrial and occupational class transformation continues. The growth of the professional and managerial middle class, many of whom work in centrally located offices, has generated both a demand for attractive housing and for proximity and accessibility to the centre. Glass (1973, p. 423) suggested that "the real risk for London is that it will become more middle class", that gentriflcation would spread across inner London and that as house prices rose they would squeeze out low-income groups. Subsequent events have largely borne out her predictions, but four things could limit or slow gentriflcation. The first is if London loses its place as the leading European financial centre. Although this currently seems unlikely, it would lead to a significant weakening of the economic base of the city and in the size and purchasing power of the middle class. The second factor is the expansion of the ethnic minority population of London and a competition for space which may ensue in some parts of inner London (see Figure 8). The 2001 Census shows that the ethnic minority population of inner London is 34 per cent and will continue to grow for demographic reasons (Scott et al, 2001; Hamnett, 2003). Although Britain does not have ghettos (Peach, 1996), the growth of ethnic minorities in inner London may limit gentriflcation. The competition is not primarily taking place in the owner-occupied sector, except in outer London where the middle-class British Asian population is growing rapidly. In inner London, the predominantly White middle class have outbid competing groups. Rather, it is taking place between the owner-occupied and what remains of the social rented and privately rented sectors and the 'battleground' is for education as much as for housing. In London, gentriflcation has almost completely transformed Notting Hill, Islington and parts of Hackney where the housing stock consists of larger Victorian terraced houses. But in other, less central or less attractive areas of the city characterised by smaller terraced houses, such as Tottenham, Brent and Clapton, minority populations are growing. It could be, therefore, that gentriflcation will reach its limit in these areas. The third factor is the growth of street crime in some areas of London (Bogan, 2000, Braid, 2000; Sengupta, 2000; Tendier and Kennedy, 2002). Middle-class households have greater freedom of residential choice than other groups and more may decide to head for the safety of the suburbs or beyond, notwithstanding commuting problems. The other thing which may limit gentriflcation is the problem of suitable educational provision for the children of the new middle class with high educational aspirations for their children and the reproduction of class advantage. While the GENTRIFICATION AND INNER LONDON 2423 60 ~] Primary -n ~J Secondary 50 rflm f 40 c § 30 Q. ill 1 rh n n ?0 10 n it Si' c 'E ■si •5 c > <ů to I O) O) >. C C .C m .B -i | g i (D a °> o o S S ra ■o 31*; »u. m O E « II "2 DC t o O) ■c 3 £ F : ■i E E t n; a) "S 3? 5 a lilii 5 ^ . C >n O) ß j3> ° » .= c 2 = g S o> mi a> Figure 8. Percentage of pupils with English as a second language, 1999. comprehensive system works reasonably well in areas with a good social mix, it can be problematic in poor areas (Gordon, 1996). Butler notes Very few of the elite ever participated in the inner London education market, and until comparatively recently, there were few middle-class families with children using inner London's schools. They were dominated by the working class—many of whom have now left. The middle class has therefore had to reconstruct an education system to meet its need to pass on inter-generational advantage to its children (Butler, 1999, p. 90). Consequently, when children reach the age of 11 and go to secondary school, middle-class gentrifiers face a fundamental dilemma. As Robson and Butler point out Among families with school-age children, there is evidence to suggest that education markets are now rivalling those in housing and employment as determinants of the nature, extent and stability of middle-class gentrification of inner-city localities (Rob-son and Butler, 2001, p. 84). This is not to lament the plight of the middle classes in inner London. There can be no doubt that, with the financial resources they command, they will always win the battle for private housing. But they cannot win the battle for educational attainment so easily. Given the continuing high percentage of social housing in inner London boroughs, the middle classes are likely to remain a minority group for years to come. Whether they will throw their weight behind the local state school system, in an attempt to improve provision, teaching and attainment, or vote with their feet or their pockets when their children reach secondary school age, remains to be seen. Conclusions The paper has argued that, while the supply of potentially gentrifiable property in the in- �8728058 2424 CHRIS HAMNETT ner areas of cities is important and needs to be explained, a key factor is the change in industrial, occupational and earnings structures that underpin middle-class demand. To this end, the paper has attempted to set out aggregate empirical evidence for these changes to show the linkages between them in London over the years from 1961 to 2001. It is clear is that the middle classes in inner London have continued to grow in both size and significance over the past 30 years. The industrial transformation of the city has been linked to a parallel occupational restructuring, linked to a long-term decline of manual working-class groups and continued expansion of professional and managerial jobs. In addition, these changes have been linked to a growth of the top end of the earnings and income distribution, leading to an increase in inequality. Not surprisingly in a market economy, the increase in the size and purchasing power of the middle classes has been accompanied by an intensification of demand pressure in the housing market. This has been particularly marked in inner London as it is here that many of the new middle class work, and this, combined with a desire to minimise commuting time, and greater ability to afford the cultural and social attractions of life in the central and inner city, has been associated with the growth of gentrification. There has also been a significant cultural and lifestyle switch which favours inner-city living which is linked to the changes in the economy, occupational class and earnings structure of London and other similar cities, such as New York, Paris and Sydney. It is because of the geographical focus of these transformations that gentrification has been strongly marked in such cities. It is a demand-based argument, but one based on major changes in industrial, occupational, earnings and income structures and related change in housing demand rather than on preferences per se. It is also argued that the transformation which has taken place in the occupational class structure of London has been associated with the gradual replacement of one class by another, rather than large-scale direct displacement. There is, however, no doubt that the price rises which this has brought about have resulted in indirect displacement in the sense that working-class residents have been priced out of most of the private housing market. References Atkinson, R. (2000a) Measuring gentrification and displacement in Greater London, Urban Studies, 37(1), pp. 149-166. Atkinson, R. (2000b) Professionalisation and displacement in Greater London, Area, 23(3), pp. 287-296. Badcock, B. (1993) Notwithstanding the exaggerated claims, residential revitalization is changing the shape of some Western cities: a response to Bourne, Urban Studies, 30(1), p. 191. Bailey, N. and Robertson, D. (1997) Housing renewal, urban policy and gentrification, Urban Studies, 34(4), pp. 561-578. Banks, R. and Scanlon, R. (2000) Major economic trends in the 1980s and 1990s: London, in: City of London Corporation, The London-New York Study. London: City of London Corporation. Berry, B. L. J. (1985) Islands of renewal in seas of decay, in: P. Peterson (Ed.) The New Urban Reality, pp. 69-96. Washington, DC, The Brookings Institution. Bogan, S. (2000) I don't know about the rest of Britain, but round here it's like the Bronx, The Independent, 1 July. Bondi, L. (1991) Gender divisions and gentrification: a critique, Transactions of the Institute of British Geographers, 16(2), pp. 190-198. Bourne, L. S. (1993) The demise of gentrification: a commentary and prospective view, Urban Studies, 14(1), pp. 95-107. Braid, M. (2000) On Ocean estate, no one except the drug gangs goes out after sundown, The Independent, 13 April. Bridge, G. (1994) Gentrification, class and residence, Environment and Planning D, 12, pp. 31-51. Brueghel, I. (1996) Gendering the polarisation debate: a comment on Hamnett's 'Social polarisation, economic restructuring and welfare state regimes, Urban Studies, 33, pp. 1431— 1439. Butler, T. (1997) Gentrification and the Middle Classes. Ashford: Ashgate. Butler, T. (1999) The new urban intermediaries? The new middle class and the remaking of London, Journal des Anthropologues, pp. 77-78. GENTRIFICATION AND INNER LONDON 2425 Butler, T. and Hamnett, C. (1994) Gentrification, class and gender, Environment and Planning D, 12, pp. 477^193. Butler, T. and Robson, G. (2001b) Social capital, gentrification and neighbourhood change in London: a comparison of three South London neighbourhoods, Urban Studies, 38(12), pp. 2145-2162. Carpenter, J. and Lees, L. (1995) Gentrification in New York, London and Paris: an international comparison, International Journal of Urban and Regional Research, 19(2), pp. 286-234. Clark, G. (2002) London in the European financial services industry: locational advantage and product complementarities, Journal of Economic Geography, 2(2), pp. 1-21. Cox, R. and Watt, P. (2002) Globalization, polarization and the informal sector: the case of paid domestic workers in London, Area, 34(1), pp. 39^17. Daly, M. (1971) Characteristics of 12 clusters of wards in Greater London. Research Report No. 13, Department of Planning and Transportation, GLC. Daniels, P. W. (2000) Business and professional services in London, in: The London-New York Study. London: City of London Corporation. Dutton, P. (2003) Leeds calling: the influence of London on the gentrification of regional cities, Urban Studies, 40(12), pp. 2557-2572. Frayn, M. (1967) Towards the End of the Morning. London: Faber and Faber. Friedmann, J. and Wolff, K. (1982) World city formation: an agenda for research and action, International Journal of Urban and Regional Research, 6(3), pp. 309-344. Frost, M. and Spence, N. (1993) Global city characteristics and central London's Employment, Urban Studies, 30(3), pp. 547-558. GLA (Greater London Authority) Economics (2002) Creativity: London's Core Business. London: GLA. Glancy, J. (1999) The other side of Shoreditch, The Guardian, 18 November. Glass, R. (1964) Introduction to London: aspects of change. Centre for Urban Studies, London (reprinted in Glass, R. (1989) Cliches of Urban Doom, pp. 132-158. Oxford: Blackwell). Glass, R. (1973) The mood of London, in: D. Donnison and D. Eversely (Eds) London: Urban Patterns, Problems and Policies. London: Heinemann (reprinted in Glass, R. (1989) Cliches of Urban Doom, pp. 159-183. Oxford: Blackwell). Gordon, I. (1996) Family structure, educational achievement and the inner city, Urban Studies, 33(3), pp. 407^124. Gordon, I. (2000) Demography, labour force and income, in: The London-New York Study. London: City of London Corporation. Hall, P. (2000) Creative cities and economic development, Urban Studies, 37, pp. 639-651. Hamnett, C. (1984) Gentrification and residential location theory: a review and assessment, in: D. T. Herbert and R. Johnston (Eds) Geography and the Urban Environment, pp. 283-322. Chichester: Wiley and Sons. Hamnett, C. (1989) Spatial and social segmentation of the London owner occupied housing market: an analysis of the fiat conversion sector, in: M. Breheny and P. Congdon (Eds) Growth and Change in a Core Region: the Case of South East England, pp. 203-218. London: Pion. Hamnett. C. (1991) The blind men and the elephant: towards a theory of gentrification, Transactions of the Institute of British Geographers, 16, pp. 173-189. Hamnett, C. (1993) The spatial impact of the British home ownership market slump, 1989— 91, Area, 25(3), pp. 217-227. Hamnett, C. (1994) Socio-economic change in London: professionalisation not polarisation, Built Environment, 20(3), pp. 192-203. Hamnett, C (1996) Social polarisation, economic restructuring and welfare state regimes, Urban Studies, 33(8), pp. 1407-1430. Hamnett, C. (1999) Winners and Losers: The Home Ownership Market in Modern Britain. London: Taylor & Francis. Hamnett, C. (2003) Unequal City: London in the Global Arena. London: Routledge. Hamnett, C. and Cross, D. (1998a) Social polarisation and inequality in London: the earnings evidence, 1979-95, Environment and Planning C, 16, pp. 659-680. Hamnett, C. and Cross, D. (1998b) Social change, social polarisation and income inequality in London, 1979-1993, Geojournal, 46, pp. 39-50. Hamnett, C. and Randolph, W. (1988) Housing and labour markets change in London: a longitudinal analysis, 1971-81, Urban Studies, 25, pp. 380-398. Hamnett, C. and Whitelegg, A. (2001) The Loft Conversion Market in Central London: A Case Study of Clerkenwell. Final report to the Economic and Social Research Council, Swindon. Hamnett, C. and Williams, P. (1980) Social change in London: a study of gentrification, The London Journal, 6(1), pp. 51-66. Harvey, D. (1974) Class monopoly rent, finance capital and the urban revolution, Regional Studies, 8, pp. 239-255. Lees, L. (1994) Gentrification in London and New York: an Atlantic gap?, Housing Studies, 9(2), pp. 199-217. Ley, D. (1980) Liberal ideology and the post-industrial city, Annals of the Association of American Geographers, 70, pp. 238-258. 2426 CHRIS HAMNETT Ley, D. (1981) Inner-city revitalization in Canada: a Vancouver case study, Canadian Geographer, 25(2), pp. 124-148. Ley, D. (1996) The New Middle Class and the Remaking of the Central City. Oxford: Oxford University Press. Marcuse, P. (1986) Abandonment, gentrification and displacement: the linkages in New York City, in N. Smith and P. Williams (Eds) Gentrification of the City, pp. 153-177. London: Unwin Hyman. May, J. (1996) Globalization and the politics of place: place and identity in an inner city London neighbourhood, Transactions of the Institute of British Geographers, 21(1), pp. 194-215. McDowell, L. (1997) The new service class: housing, consumption and lifestyle among London bankers in the 1990s, Environment and Planning A, 29, pp. 2061-2078. Merrett, S. (1988) The Denarius hypothesis. Greater London Housing Study Working Paper, Bartlett School of Architecture and Planning, University of London. Mollenkopf, J. and Castells, M. (1991) Dual City: Restructuring New York. New York: Russell Sage Foundation. Munt, I. (1985) Economic restructuring, culture and gentrification: a case study of Battersea, London, Environment and Planning A, 19, pp. 1175-1197. Pahl, R. (1988) Some remarks on informal work, social polarisation and the social structure, International Journal of Urban and Regional Research, 12, pp. 247-267. Peach, C. (1996) Does Britain have ghettos?, Transactions, Institute of British Geographers, 21, pp. 216-235. Peach, C. (1997) Pluralist and assimilationist models of ethnic settlement in London, 1991, Tijdschrift voor Economische en Sociale Geografie, 88, pp. 120-134. Pratt, A. (1997) The cultural industries production system: a case study of employment change in Britain, 1984-91, Environment and Planning A, 29, pp. 1953-1974. Preteceille, E. (1995) Division sociale de ľespace et globalisation: le case de la metropole parisienne, Societies Contempo-raines, 22/23, pp. 33-67. Redfern, P. (1997) A new look at gentrification: gentrification and domestic technologies, Environment and Planning A, 29, pp. 1275-1296. Robson, G. and Butler, T. (2001) Coming to terms with London: middle class communities in a global city, International Journal of Urban and Regional Research, 25(1), pp. 70-86. Sassen, S. (1991) The Global City: New York, London and Tokyo. Princeton, NJ: Princeton University Press. Savitch, V. (1988) Post-industrial Cities: Politics and Planning in New York, Paris and London. Princeton, NJ: Princeton University Press. Schaffer, R. and Smith, N. (1986) The Gentrification of Harlem? Annals of the Association of American Geographers, 76(3), pp. 347-365. Scott, A. J. (1997) The cultural economy of cities, Journal of Urban and Regional Research, 21, pp. 323-339. Scott, A., Pearce, D. and Goldblatt, P. (2001) The sizes and characteristics of the minority ethnic populations of Great Britain - latest estimates, Population Trends, 105, pp. 6-15. Sengupta, K. (2000) A terrible place to die. A terrible place to live, The Independent, 30 November. Simmie, J. (1983) Beyond the industrial city, Journal of American Planning Association, Winter, pp. 59-76. Smith, N. (1979) Toward a theory of gentrification: a back to the city movement by capital not people, Journal of the American Planning Association, 45(4), pp. 538-548. Smith, N. (1987) Of yuppies and housing: gentrification, social restructuring and the urban dream, Environment and Planning D, 5, pp. 151-172. Smith, N. (1996) The New Urban Frontier: Gentrification and the Revanchist City. London: Routledge. Smith, N. and deFilippis, J. (1999) The reasser-tion of economics: 1990s gentrification in Lower East Side, International Journal Urban and Regional Research, 23(4), pp. 638-653. Smith, N. and Williams, P. (1986) Gentrification of the City. London: Allen & Unwin. Tendler, S. and Kennedy, D. (2002) London faces eruption of ethnic crime wars, The Times, 21 November. Thrift, N. and Leyshon, A. (1992) In the wake of money: London and the accumulation of value, in: L. Budd and S. Whimster (Eds) Global Finance and Urban Living, pp. 282-311. London: Routledge. Warde, A. (1991) Gentrification as consumption: issues of class and gender, Society and Space, 9, pp. 223-232. Williams, P. (1976) The role of institutions in the inner London housing market: the case of Islington, Transactions of the Institute of British Geographers, NS, 1(1), pp. 72-82. Williams, P. (1978) Building societies and the inner city, Transactions of the Institute of British Geographers, NS, 3, pp. 23-34. Wyly, E. K. and Hammel, D. J. (1999) Islands of decay in seas of renewal: housing policy and the resurgence of Gentrification, Housing Policy Debate, 10(4), pp. 711-771. Copyright of Urban Studies is the property of Carfax Publishing Company and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.