John K. Walton

Seaside resorts at the turn of the century
At the beginning of the twentieth century the English and Welsh coastlines were uniquely well-endowed with seaside resorts, as befitted a culture which invented the commercialization of sea-bathing and the distinctive and often highly-specialized settlements which served the fashion and its offshoots. There were clusters of large centres and their satellites in particularly attractive and accessible locations, but every coastline had its resort towns and villages, in a bewildering variety of types and catering for an almost infinite range of social groups and cultural preferences. This was by now a well established resort system, or group of systems, with a history of growth and change which went back to the eighteenth century. The resort map of England and Wales as the Edwardians saw it owed more to the Victorian years and the maps of demand which railways, population concentrations, changing income levels and social structures, and new fashion in the pursuit of health and pleasure, had helped to draw during this dynamic and formative period. European competitors for well-off British holidaymakers had emerged, especially on the French coasts, but there was plenty of growth in demand at working-class and mainstream middle-class levels to sustain continued expansion beyond the turn of the century (Walton, 1983; Travis, 1993).

The years between the 1881 and 1911 censuses, and especially the decade of the 1890s, saw the fastest growth in English seaside resort populations, as new competitors appeared for established resorts which themselves often continued to expand impressively. During these 30 years the aggregate off-season population of 145 English and Welsh seaside resorts grew by more than 600,000 to just over 1.6 million — nearly 4.5 percent of the total population (Walton, 1983; Pimlott, 1947). Fourteen seaside resorts more than trebled their census populations over this period, including Blackpool, Southend and Bournemouth, while a further 22 more than doubled in size on this measure (Walton, 1983). Such growth-rates were not quite sustainable in the new century, as we shall see, but it is important to emphasize at the outset the scale and rapidity of resort growth at the beginning of the period, which set the tone for future developments in many ways.

The surge of growth which ushered in the new century was unprecedented. It was fuelled above all by summer holidaymakers, as the week, fortnight or even month at the coast became more generally attainable across the middle strata of society, while professional, managerial, administrative and other salaried jobs became more numerous. Meanwhile, in some parts of the country working-class demand was beginning to make its presence felt, especially on the coasts of Lancashire and Yorkshire and within reach of London. As resorts grew and proliferated, more of them became heavily dependant on a three-month summer season, and this boosted the already high levels of mobility into and out of these rapidly growing settlements. The accommodation, transport, leisure and other service industries attracted seasonal workers as well as permanent residents, and the fragility in practice of enticing-looking business opportunities at the seaside made for high turnover levels and generated migration into and out of the resorts. So residence at the seaside often lasted for only part of the year, for workers as well as visitors, and even for year-round residents their stay in a resort was often a life-cycle stage. Young men tended to leave in search of better opportunities elsewhere, while women were attracted as domestic servants in their teens and twenties, and as lodging-house keepers later in life. Retirement to the seaside was also beginning to make a difference to seaside resort demography in certain areas, especially Sussex and Devon and, long before retiring age, businessmen and families of independent means were choosing to live in coastal locations at a considerable distance from their businesses or places of origin. All this made for highly volatile resort populations, and we shall explore some of the consequences for seaside society later. Meanwhile, the point to emphasize here is that these high levels of mobility meant that, in practice, seaside life was much more widely experienced than even the impressive snapshot census figures of resort populations would suggest, The swelling flood of summer visitors adds additional force to this argument. So seaside resorts were becoming even more important than the census population figures make them look.

The pattern of resort distribution in the early twentieth century owed much to the legacy of early development, refracted through the changes of the railway age. More than 40 per cent of the seaside resort population in 1911 lived in Sussex, Kent and Hampshire, the pioneer areas for seaside resort growth. Most lived in Sussex, five of whose coastal resorts featured among the 10 fastest-growing seaside watering-places in absolute terms between 1881 and 1911. so the areas where seaside tourism had begun, in South-east England, were still both prominent and dynamic at the start of the new century. The most impressive challenge to this (by now) traditional dominance came from Lancashire and North Wales, where an added dynamism was generated in established mid-Victorian resort economies by the extra ingredient of working-class demand for full-scale holidays (as well as day trips) from the factory towns of Lancashire and Yorkshire and from the West Midlands. This development was fuelled especially by the textile centres of Lancashire and the West Riding of Yorkshire, and above all by the distinctive and relatively prosperous working-class family economies of the Lancashire 'cotton towns'. The Lancashire coastal resorts, indeed, showed a spectacular combination of high growth rates in population and visitor numbers across the turn of the century, with Blackpool taking the lead in both respects (Walton, 1981, 1992). The Essex coast was also coming into the running, fed by London demand, but there was stagnation as well as expansion at the seaside during these years. At one extreme, Blackpool, Bournemouth and Southend, with their contrasting markets and social mixtures, were prominent among the fastest-growing English towns at the turn of the century. On the other hand, more low-key versions of the seaside holiday trade buttressed the sagging economies of declining ports and centres of maritime industry, while there were many stagnating fishing villages whose languishing economies were helped out by the summer contributions of self-consciously discriminating and quiet-seeking visitors from the cultured professional and literary middle classes. .1n neither case did much in the way of population growth result: often it was more a matter of mitigated decline (Waller, 1983).

Themes in resort development, 1900—1950
How to generalize across the intervening layers of experience, the kaleidoscopic variety of resort sizes, economic structures, aesthetics, visiting publics and levels of sophistication, is a problem which will recur. It will repay the effort, because the seaside resort was and is an important kind of town. As well as contributing to a broader understanding of English and Welsh society in this period, an analysis of seaside resort development will promote an appreciation of the importance of inherited characteristics and enduring traditions in helping to explain the current problems of the resorts themselves. Moreover, seaside resorts and the holiday industry provided employment and outlets for capital investment on a scale which is still not generally realized, as well as constituting distinctive kinds of environment and social system which affected far more lives than those whose presence at the seaside was captured by the decennial censuses. The seaside also formed — and continued to be — a special case with regard to employment patterns and social structure' With its own characteristic problems of seasonality, irregular employment' child labour, gender relations in the family economy, and the poverty and special needs of increasingly well-represented elderly elements in the population. Seaside resorts were often crucibles of conflict over lifestyles, as visitors from differing and sometimes incompatible backgrounds with divergent expectations about holiday enjoyments and environments competed for access to and use of valued space within resorts, although by the early twentieth century sophisticated patterns of social segregation between and within resorts had emerged to mitigate these tensions. However, retired residents and people of independent means, along with a broadening stratum of long distance business commuters and commercial travellers in accessible resorts with improving train services, added a further dimension to the persisting debates over 'social tone', especially as such people tended to be interested in low municipal taxation rather than the provision of services and amenities to visitors by local government, or expenditure on advertising and publicity. The role of local government in resort economies is a particularly strong theme, as it expanded its role through the period at the expense of private companies and landed estates. Many of its activities were innovative, as councils spent heavily on parks, bathing pools, solaria and sporting facilities to meet new visitor expectations, while road improvements and car and coach parks marked the municipal response to the internal combustion engine. New developments in what resorts had to offer constitute a general theme of great significance. The experience of seaside holidays was sufficiently widely shared by the early twentieth century for traditions to have been invented and enshrined in a popular folklore of the seaside, finding expression in the humour of the comic postcard and the popular press, as well as in the style and content of distinctive seaside entertainments themselves, from bathing machines to donkeys to pierrots. The seaside was already making its own contribution to popular culture, and It discarded at its peril; but during the inter-war years especially, new styles of holidaymaking were being grafted on alongside the old, along with the changing fashions in clothing and building styles which marked these years. The ways in which the new was incorporated, and the extent to which it needed to find its satisfactions in new places, need to be addressed. There is also a political dimension. Seaside holidays were often seen as passive, reactive enjoyments which reinforced political quiescence and acceptance of the status quo, in contrast with the radical activism associated in some minds with an emergent 'outdoor movement' of hikers and touring cyclists. The larger resorts, at least, were certainly political bastions of the Conservative Party, but how far they exercised a more generally conservative influence on residents and visitors is a matter for debate, just as the political complexion of hiking, rambling, cycling and climbing should not be over-simplified. Pulling the themes of this paragraph together will provide an extensive agenda for discussion, not all of which can be pursued in this chapter (see Taylor, 1993).

Further analysis will need to grow out of a perspective on the pattern of resort development during the first half of the twentieth century. Which were the most dynamic resorts and coastlines, and where was stagnation or decline to be found? What kinds of resort prospered during this period, and why? The limited existing historiography on this period is not very helpful here, and systematic treatment is best approached through the census figures on resort populations, with all their limitations (Walvin, 1978; Pimlott 1947; Brunner, 1945). We lack any remotely plausible evidence on visitor numbers, which would need to be pursued through a range of local sources and which be unlikely to be at all reliable.

Several problems obtrude immediately when we approach the census data for these purposes. In the first place, two censuses are effectively missing: that of 1941 was never taken because of the Second World War, and that of 1921 was taken in late June, when the season was much more strongly developed than at the usual time in early April. Attempts were made to adjust for the obvious boost to resort populations which the advent of the season had brought about, but the results were unconvincing. We can learn something about early-season visitor numbers from the 1921 census, but it is best omitted from attempts to establish population trends (Walton and O'Neill, 1993). Then we encounter the problem of how to define a seaside resort for these purposes, given that so many places combined resort with other functions including manufacturing industry as well as fishing and seaport ties. Any attempts to draw precise lines between seaside resorts and other kinds of place would require detailed specifications about social structural and other characteristics which would ultimately be based on arbitrary criteria. Moreover, the balance between different elements in urban economies may change over time, pulling particular places into and out of side resort status. The case of Fleetwood, where the local authority promoted the holiday industry when trawling was in the doldrums, only to let matters slide when the fishing picked up, is an instructive illustration (Hutchinson, 1973). This is a reminder that raw census totals tell us nothing about health of the resort element in a complex local economy, as at Whitby, where overall population stagnation arose from the combination of a lively holiday trade and declining maritime and manufacturing activity (Horne and Son Ltd, 1934). A relate problem of disaggregating resorts from the larger census districts of which they may form part. Southsea's entanglement with Portsmouth is the most important example, but others include the Roker area of Sunderland, the New Brighton area of Wallasey, and on a smaller scale the resort of Silloth within the larger administrative area of Holme Cultram in Cumberland. Small resorts are just as likely to pose problems here as important resort districts within large cities, and the inclusion of such places within a list of resorts for the purpose of comparison between censuses would require extensive local research into the county reports and the boundaries of resort areas, which lies beyond the scope of this chapter. Local knowledge is also required to take full account of boundary changes, and some of figures which follow may be distorted by failures to adjust for these.

Nevertheless, some broad indications of trends can be found by following a sample of 116 places with strong pretensions to seaside resort status through the censuses (see Figure 2.1 and Table 2.1). Inclusion in the list is based on reputation and on advertising visibility and promotional activity rather than on more demanding 'scientific' criteria, and places which pose serious problems of boundary definition and identity have been omitted. A full roll-call of seaside resorts in England and Wales, including villages whose development had passed little beyond the 'discovery' stage, would count perhaps twice as many. The focus is on substantial settlements which had reached, or were to reach, some degree of maturity. The main thrust of what follows is to compare the picture in 1911 with that in 1951, taking some account of the 1931 evidence where changes in trend seem to have occurred. We begin with some general contextual remarks.

In the first place, seaside resorts as a kind of town continued to grow throughout the period, apart from the interruptions of the two world wars. In some cases growth was accompanied by an increasing specialization in retirement or dormitory functions at the expense of the holiday industry as such; but even so, the overall population trends are impressive, especially if we bear in mind growing competition from the Continent and cruise liners at the top end of the market, the rise of new forms of domestic holidaymaking and the prevalence of trade depression in some areas in the inter-war years. Moreover, this continuing growth, in old resorts as well as new ones, took place at a time when many manufacturing towns with similar eighteenth- and nineteenth century origins were stagnating or declining (General Register Office, 1950). Between 1911 and 1931 the seaside resort population grew by about 25 per cent, from 1.6 to 2 million; and over the next 20 years, which also included a world war, it increased by a similar order of magnitude to nearly 2.5 million. The 116 resorts in the sample, which included almost all the sizeable ones, increased their populations by 50.4 per cent. This meant that seaside resorts actually increased their share of the population of England and Wales, from nearly 4.5 per cent in 1911 to 5 per cent in 1931 and 5.7 per cent in 1951. They did not sustain their erstwhile leading role in urban development during these years, but the most dynamic of them did feature among the fastest-growing towns, among lists which were dominated by London suburbs (with a few dormitories elsewhere thrown in), new manufacturing towns and expanding older settlements in the South Midlands, and a few new northern industrial centres like Scunthorpe and Billingham. Between 1921 and 1931 (after the notional deduction had been made for 'visitors' in the 1921 count) Blackpool was eighth, Poole (which had resort as well as more conventional industrial functions) ninth, Southend tenth, and Bournemouth twelfth among the 28 'large towns' (with populations greater than 50,000 in 1931) which had grown by more than 10 per cent over the decade. The top seven places in this list went to London suburbs, further emphasizing the residential commuter/retirement nature of most urban expansion in this period (Jackson, 1973). A similar mix of kinds of town featured among the 49 smaller places whose populations had grown by more than 5000 during the same decade. Thirty-three were outer London suburbs, but Paignton, Torquay, Morecambe and Worthing also made a strong showing, and several others were only just outside the frame. Moreover, the pace of resort growth over these 40 years was not markedly inferior to the peak decades of 1881—1911, if we bear in mind that 10 years between 1911 and 1951 saw the disruptions of war, making the second period more comparable with the first than might originally appear. If we look at percentage growth rates, 10 resorts trebled their population (at least) between 1911 and 1951 compared with 14 for the earlier period, while a further 22 (as in 1881—1911) saw their inhabitants double in number. In the later period, of course, more resorts were starting from a relatively high base figure; and if we compare incremental growth we find that the top 10 resorts in this respect added 412,943 to their populations during 1911—1951 as against 310,317 during 1881—1911. So at the top end of the scale seaside resort populations were actually accumulating more rapidly during the later period, boosted by gains of 86,708 for Blackpool and 89,093 for Southend which far outscored the 60,067 which put Bournemouth at the top of the corresponding table in 1881—1911.

This leads into discussion of the size-structure of the resorts in the sample (which, we must remember, leaves out a lot of the smaller places, as well as Southsea and some other resort districts of large towns). Contrary to the expectations which might be raised by the incremental growth figures, the relative importance of the largest resorts declined somewhat between 1911 and 1951, partly because some of them actually lost population (Dover and Great Yarmouth stand out here) while others (like Eastbourne and Hastings) stagnated. In 1911 the 10 largest resorts accounted for 44 per cent of the population, but 40 years later their share had declined to 40.9 per cent. The corresponding figures for the top 20 were 62.7 per cent and 57.7 per cent. This outcome reflects a lot of impressive growth among the middle-ranking resorts, and we shall see that many of the most dynamic performances in percentage terms came from within this group.

It is, indeed, time to move on to discuss which coastlines and resorts fared best in population growth terms during this period, and to try to assess the reasons for and significance of the patterns which emerge. In terms of sheer weight of numbers, growth was dominated by a group of substantial resorts which contrived to sustain their dynamism into the new century; but a different and very interesting pattern is revealed by the percentage growth figures from the sample. We begin with the incremental growth evidence.

The most substantial incremental population growth in English and Welsh seaside resorts during 1911—1951 was concentrated overwhelmingly in large, established towns in regions with a long history of seaside tourism. Blackpool, Southend and Bournemouth, which had drawn away from the pack during 1881—1911, sustained their dynamism to continue to extend their lead at the top, although Brighton grew sufficiently to keep its position as the most populous English resort, and its neighbour Hove was just behind in the table. These old-established centres occupied fifth and sixth places in the incremental growth table, with nearby Worthing moving into fourth place after having been ninth in the earlier period: in 1911 it had been the fifteenth most populous resort in the sample. Six others among the 20 fastest-growing resorts on this definition during 1911—1951 had also been in the top 20 between 1881 and 1911. Of these, Southport, Torquay, Folkestone and Weston-super-Mare had also featured among the 20 most populous resorts at the 1911 census, while Hoylake and Whitley Bay had been just outside the frame. Of the rest, Margate and Weymouth were resorts with very distinguished pasts which had recovered well from a period of relative stagnation in the Victorian years, but the others were advancing rapidly from humbler and more recent origins. Bognor Regis, Morecambe and Paignton had been relatively late developers on highly competitive coastlines, while Clacton was a beneficiary of the channelling of metropolitan white-collar demand to the Essex coast. Redcar's growth was boosted by industry, as an offshoot of Teesside, and Christchurch, which contained three small resorts within its boundaries, was also in some respects an extension of Bournemouth. There was little that was really novel in this pattern: the established resort areas were reaping the advantages of their accumulated attractions and power to advertise, and challenges to their dominance arose from neighbours which were developing distinctive identities of their own, rather than from new coastlines. Compared with the preceding period there was relatively little spectacular growth from small origins: at most, as in the case of Bognor or Morecambe, resorts were moving in on the big league from already well-established positions a little lower down the scale.

Not all the established large resorts sustained earlier large-scale growth in the new century, however. The decline of Dover, which had been growing strongly at the turn of the century, was the most remarkable reversal. Eastbourne lost the momentum it had sustained during the late Victorian and Edwardian years, and so did nearby Hastings, while Bexhill, between the two, continued to grow steadily without maintaining its very rapid growth of the years between 1881 and 1911. The growth of Penarth and Lowestoft was also checked sharply, as was that of Cleethorpes and Colwyn Bay. In all these cases smaller resorts nearby took up the running, at the expense of their older rivals. Significantly, however, what seems to have been happening — and there are other examples, such as Ramsgate and its neighbours on the Kent coast — is a realignment of demand and investment within existing resort systems, rather than the dramatic emergence of new coastlines. The much advertised resorts of Devon and the 'Cornish Riviera' failed to deliver large-scale growth outside the Torbay area, which had deep roots as a resort coastline and had sustained growth over a long period.

Bungalow towns and holiday camps
When we take the alternative perspective of analysing percentage population growth, however, a much clearer pattern emerges among the places which were making rapid headway from small beginnings. The highest levels of the resort hierarchy may have been self-perpetuating, but there was a lot of turbulence lower down the scale. Almost all of the 30 fastest-growing resorts in percentage terms were boosted by either informal plotland developments of 'do-it-yourself' bungalows and adapted railway carriages or tramcars, or by the pioneer holiday camps of the 1930s and immediate post-war years. The list excludes a string of smaller places, from the infamous Peacehaven development on the Downs near Brighton (with a population of just over 3000 in 1951) down to scatterings of huts and caravans on various stretches of coastline. However, some places in the sample embraced both kinds of settlements, and the evidence linking rapid seaside population growth with this kind of informal, Bohemian provision for people of limited means is overwhelming.

Standing out at the top of the list were Canvey Island, a remarkable story of rapid growth which would have been equally impressive in Victorian times, and the Sussex resort of Lancing, hitherto notable mainly for its public school. Fittingly, Canvey Island was the ultimate plotland development. Its agricultural land had collapsed in value during the late Victorian depression, which hit Essex particularly hard, and in 1899 a Southend land agent bought up an extensive tract of the eastern half of the Island and offered it for sale in small lots to people who wanted cheap, informal holiday homes. The original promoter, Frederick Hester, went bankrupt within a few years, and his grand design for a 6-mile linear Winter Garden along a canal was never realized, but the proliferation of shacks and bungalows continued, as Canvey offered a seaside retreat to people from the East End of London who could never otherwise have afforded it. An oil storage depot was developed at the western side of the island from 1936, in the teeth of fierce opposition from the locals; but Canvey Island remained overwhelmingly a holiday place, with growing numbers of year-round residents and (by 1951) the elderly. By this time its population had been boosted by refugees from the Blitz and the post-war London housing shortage. This was the new seaside resort experience, loathed by planners and aesthetes and loved by its proud householders and relaxed visitors, in excelsis (Cracknell, 1959; Hardy and Ward, 1984). Such schemes did not always 'take': a similar one at Minster-in-Sheppey, on the other side of the Thames Estuary, languished in spite of the promotional efforts of another Southend entrepreneur, Frederick Ramuz, although it provided cheap and accessible retreats for the occupants of makeshift but enduring chalets scattered through the plots (Hardy and Ward, 1984; Pahl, 1984).

This kind of provision for clerks and artisans (and for the Bohemian middle classes of artistic bent and theatrical associations) was spreading in more piecemeal ways across many coastlines between the 1890s and the inter-war years, usually owing more to small local enterprise than to large-scale capitalistic land speculators. Lancing was a (characteristically) more complicated example than Canvey Island. It had market gardens and (from 1912) a railway carriage works as well as an established visiting season; and between 1890 and 1905 about 200 'bungalow town' houses were put up on a shingle spit in front of the village. A storm in 1913 disposed of many of them, but building resumed along the coast road in the inter-war years, and Lancing acquired an urban aspect, with more conventional bungalows intermingled with houses for retirers and Southern Electric commuters. There were still three caravan parks in the early 1950s, and the holiday season had not quite given way to residential specialization; but the path in this latter direction was well-marked and widely followed (Hudson, 1980; King, 1984).

At least 13 other resorts among the top 30 in the percentage growth table owed some of their growth to this kind of cheap, informal, ad hoc development. For example, Mablethorpe in Lincolnshire fits well into this pattern. Prestatyn and Abergele, which also followed hard on the heels of Lancing, were part of a length of North Wales coastline between Prestatyn itself and Llandudno which proved hospitable to chalet and bungalow development in these years (King, 1984). This included a particularly ambitious venture at Kinmel Bay, near Rhyl, and this latter resort, already well-established at the turn of the century, also made a strong showing in the percentage growth league for 1911—1951. Clacton, with its Jaywick Sands development (whose street names celebrated makes of motor car, whose associations with freedom and spontaneity overlapped with the preferred image of this kind of settlement), was part of an Essex coast variant on this theme. Older Sussex resorts which contained 'bungalow town' enclaves similar to that of Lancing, and featured in the top 30 in the sample, were Bognor Regis, Shoreham and Worthing, although in these cases, as in that of Blackpool, the actual contribution of such initiatives to population growth was marginal, although it should be stressed that the 'bungalow towns' soon acquired year-round populations, attracted by their cheapness as well as their location and 'backto-nature' attributes. The key point, perhaps, is that many of the fastestgrowing resorts in percentage terms were 'open' and unrestricted enough to find room for such development (suspicious though local authorities might be in such resorts as Clacton), and the free-and-easy approach to holidays that went with it.

Holiday camps also partook of this ethos, offering an alternative to the formality of hotels and the restrictive regime of the seaside landlady. A holiday camp might not boost a resort's population at an April census, but it represented an approach to holidaymaking which seems to have become widely popular, especially in the inter-war years, and not just among the small minority who patronized the camps themselves. At least 17 of the 30 fastestgrowing resorts in the percentage table had holiday camps of one sort or another, and some, especially in North Wales, Essex and Lincolnshire, had several. Ward and Hardy's map of the holiday camps of England and Wales in 1939 is in large part a map of the fastest-growing seaside resorts of the preceding generation (Ward and Hardy, 1986). At least nine of the 30 resorts had both bungalow or chalet settlements and holiday camps, including Canvey Island itself, Prestatyn, Abergele, Bognor Regis, Clacton, Blackpool, Whitstable, Withernsea and Rhyl.

We shall see that these were not the only relevant themes, but they stand out so clearly that they deserve to be highlighted here. Of the resorts in the top 30 which seem to have had neither of these distinguishing features (and there are only seven in all), Thornton/Cleveleys and Redcar combined extensive industrial activity with their resort functions, and Hoylake was as much a Merseyside commuter settlement as a health and golfing resort. Felixstowe already combined port and resort functions. The point is that there were other ways of arriving in the ranks of fastest-growing resorts between 1911 and 1951, but association with informal development and holiday camps was the most widespread and obvious.

The West Country and East Anglia
Some of the absences from leading positions in the percentage growth table merit comment. We might expect the rising resorts of Devon and Cornwall, and perhaps of the Norfolk coast, to feature here, even if they could not muster the sheer weight of additional numbers to show up in the incremental growth lists, but their only representatives in the top 30 are Paignton, Bude and Newquay. Three of the more up-market south Devon resorts, Budleigh Salterton, Seaton and Sidmouth, grew at more than the rate for the whole sample, but this was not an impressive showing if we consider the advertising resources which were devoted to promoting this region by (especially) the Great Western Railway (Smith, 1988). Norfolk, meanwhile, was not a growth area: Cromer, Sheringham, Hunstanton and Mundesley stagnated or showed very limited expansion, failing to follow up their impressive development during 1881—1911, while Great Yarmouth's population actually declined, although this does not mean that its holiday industry was itself in bad shape.

Some small but relevant growth areas in Norfolk have not been picked up by the sample, including holiday camp and chalet developments at places like Caister and Heacham, but this will not have made much difference to the outcome. More might be made of this theme in the South-west, however, and especially in Cornwall. After all, by 1951 the South-west region was said to be attracting 14 per cent of 'main British holidays', although the main growth was to come later, as the corresponding figure for 1973 of 23 per cent of a much larger domestic holiday market indicates (Morgan, 1992). More headway was certainly being made than the bare population figures would suggest. In Devon, Morgan (1992) suggests that inter-war tourism remained concentrated into and adjacent to the established rail-served resorts, despite the increasing importance of motor-coach tours from London in these years. There were also pressures to develop holiday camps, some of which came to fruition despite opposition from countryside campaigners and the existing holiday industry. In 1944, however, there were only seven sizeable camps in the county, three of them at Paignton, which had a more popular clientele than the other South Devon resorts and showed the fastest percentage population growth. There were still only 1037 caravans in Devon in 1949, though their numbers were to multiply 10-fold in the next nine years; and it was the 1950s that saw the great camping and caravan boom in the county. Meanwhile, small resorts outside the sample, such as Combe Martin and Woolacombe, were attracting a lot of visitors in proportion to their resident populations by the 1930s, while Brixham, whose population growth was very limited, more than doubled its boarding-house numbers during the 1930s, as the base of its economy shifted from trawling to the holiday trades (Morgan, 1992). Where there is slack in the economy as older occupations decline, of course, growth in the holiday industry can be achieved without making much impact on population expansion. This was the case in parts of Devon, and even more so in the coastal districts of Cornwall.

Shaw and Williams show that Cornwall's coastal parishes declined slightly in population during 1911—1921, but increased by 3.2 per cent during 1921—1931 (after deducting the visitors in the former year) and then by 10 per cent during 1931—1951 (Shaw and Williams, 1991). This accelerating but limited growth is in keeping with the findings for the larger resorts, where only Newquay really stood out on the national stage. However, this does not take full account of the emergence of pockets of holiday accommodation and services away from the towns and villages, which was an important inter-war theme. Such developments took the form of single large hotels with attendant shops, or estates of holiday homes, or less formal scatterings of isolated bungalows, or chalet, caravan and camping sites. Such innovations were associated with retirement and with more down-market holidaymakers, as well as with more affluent motorists, and their growth accelerated rapidly in the immediate post-Second World War years. Moreover, in Cornwall as in Devon, the holiday trades grew by filling the gaps in local economies and counteracting out-migration rather than by stimulating much visible growth, although this tendency was masked by changes in the way the census recorded the key service industries (Shaw and Williams, 1991; Perry, 1993 — many buildings of this kind would not be occupied at the time of the census, here as elsewhere).

The growth of seaside resorts and holidaymaking in the South-west was thus more impressive than the bare population figures make it look. Even so, as Morgan points out for Devon, inter-war tourism remained concentrated into and adjacent to the traditional railway-served resorts for the most part: the new growths were a minority preserve (Morgan, 1992). Williams and Shaw (1993) argue that by 1936 'motor transport had left an indelible mark on the tourist industry' in Cornwall, with Penzance being promoted as a touring centre and road improvements in progress; but in 1952 there was still only one set of traffic lights in the whole county (perhaps significantly, at Newquay) and planners were anxious to improve road access quickly to cash in on post-war holidaymaking. More generally, it seems that resort growth nationally during 1911—1951 was concentrated mainly in existing strongholds, with established patterns of holidaymaking still predominating in many ways and new growths and distant coastlines remaining the preserves of growing but still marginal minorities. The disproportionate concentration of resort growth in Devon into Torbay is another regional illustration of this trend.

Prosperity and accessibility
This general impression is borne out when we look at changes in the resort population rankings in the sample. Those which rose by 10 or more places, indicating comparative success in broader context, were those that dominated the percentage growth rankings — the leaders in terms of aggregate growth were already too near to the top to figure on this list. Those which fell by 10 or more places were declining seaports with a holiday industry, with a scattering of more conventional specialized resorts like Penarth, Aberystwyth, Ventnor, Cromer, Tenby and Lyme Regis, some of which had been expanding rapidly before 1911. All were on the tourism periphery, away from the most popular and accessible, and the most developed, coastlines. Eight out of the 25 were in Devon and Cornwall, including Ilfracombe, which was also part of a Bristol Channel holiday economy which linked it with South Wales by steamer. On the Welsh side of the water Penarth and Tenby were also in comparative decline. The rise of Porthcawl and Barry Island, coupled with the growing attractions of Weston-super-Mare and Burnham and the sustained depression in the staple South Wales industries, must have contributed to this pattern. Other resorts which were failing to compete successfully were concentrated in East Anglia, North-east England and West Wales, along with the two North Wales resorts which (apart from Anglesey) were furthest from the main industrial centres of Northern and Midland England, and suffered competition from the more accessible Prestatyn, Abergele and Rhyl. Llanfairfechan and Penmaenmawr hardly grew at all during these years. Two special cases were Dover, which declined in population even as it pushed its holiday attractions at French visitors (the Dover Holiday Guide, 1935, had a German-language supplement as well as a French one), and Ventnor, which as in part a climatic resort, probably suffered especially from overseas competition but the general trend is clear.

Five seaside resorts were sufficiently expansive to appear prominently in all three measures of growth: the incremental, the percentage and the changing population rankings. These were Morecambe, Bognor Regis, Redcar, Christchurch and Clacton. Redcar is a special case because of its industrial growth, but it did claim 2 million visitors in 1933, most of whom must have been day-trippers from Teesside, and it may have prospered at the expense of nearby Saltburn and Whitby (Dougill, 1935). The other resorts share a great deal in common with other expanding resorts in this period, and they may well be exemplary cases of a wider pattern. Before exploring this further, something must be said about visitor numbers and profiles.

The imperfections of census populations as measures of resort success have already been emphasized, but it is very difficult to assemble plausible evidence on visitor numbers for this period, still less on length of stay and spending power. Estimates for the mid-1930s from the resorts themselves gave Blackpool 7 million visitors per year, Southend 5.5 million, Hastings 3 million, Rhyl 2.5 million, Bournemouth, Southport and Redcar 2 million, and Eastbourne and Morecambe over 1 million (Brunner, 1945); but there were guesstimates — there was no mechanism for arriving at accurate figures, and (for example) it seems on the face of it unlikely that Rhyl had two-and-a-half times as many visitors as Morecambe, or that Southport could have matched Bournemouth. Something can perhaps be done with railway statistics: thus passenger traffic through the South Devon resorts on the Great Western Railway increased by 65 per cent between 1913 and 1923, only to fall again after 1929 to reach 71 per cent of the 1923 level by 1934 (Morgan, 1992); but it is impossible to disaggregate holidaymakers from other kinds of travellers from this evidence, or to allow for the growing importance of road transport.

We can also obtain some clues from the censuses of 1921 and 1951. From the June census of 1921 the Registrar-General's office calculated the number of visitors in each resort on 19 June and expressed it as a percentage of the 'estimated resident population'. This gives us a comparative snapshot of the popularity of resorts just as the main season was about to begin, although some of the estimates were angrily disputed in the resorts themselves. In terms of sheer visitor numbers it was the big popular resorts that led the field, even though their main harvest in July and especially August still lay in the future. The top three were Blackpool, Margate and Southend. Lower down the top 20 resorts, what stands out on this basis is the prominence of middle-sized resorts catering for the upper working class and lower middle class, such as Bridlington, Weston-super-Mare, Ramsgate, Skegness, and the inevitable Clacton and Morecambe (though not Bognor Regis). A few select resorts showed strongly, such as Bournemouth, Eastbourne and Torquay, although they were already diversifying their markets downwards; but it was the popular resorts that carried the big guns. In percentage terms, the picture was rather different, although the two tables held seven names in common among the middle-sized popular resorts: Skegness, Margate, Bridlington, Llandudno, Clacton, St Anne's and Morecambe. The different emphasis came from a clutch of small- to medium-sized resorts which provided quiet holidays for middle-class families, such as Barmouth, Shanklin, Lynton and Frinton. One interesting point to emerge is that both ways of counting the most attractive resorts included several which were not growing much in terms of their residential populations. Among the resorts with the largest numbers of visitors, Scarborough was fifth, Bournemouth sixth, Hastings ninth and Ramsgate eleventh. Margate, which showed below-average growth between 1911 and 1951, was second in the visitor numbers table and fourth in the percentages table, where small resorts with stagnating resident populations were prominent. Barmouth was second, Shanklin third, Lynton sixth, Grange-overSands fifteenth and New Hunstanton sixteenth. These were small but perhaps particularly highly-specialized resorts with distinctive attractions (Walton and O'Neill, 1993).

The 1951 census offered a table showing the number of people enumerated on census night in each district, but resident elsewhere in England and Wales or further afield. This gives an interesting view of the comparative importance of a spring season, with visitors captured at the beginning of April; and here small resorts which had exhibited very slow population growth predominated. The list was headed by Lynton in North Devon, where 16.6 per cent of those counted came from elsewhere in England and Wales. Others with visitors accounting for more than 10 per cent of the population were (in order) Grange-over-Sands, Ventnor, Lyme Regis, and Salcombe in South Devon, with Sidmouth, Towyn, Swanage, Aldeburgh and Broadstairs making up the top 10. This is a very different list from those encountered hitherto, being dominated by select resorts with ascribed climatic advantages which had in some cases chosen to remain small, as local authorities and landowners restricted development and discouraged amusements (General Register Office, 1956). The 1951 evidence suggests that they also benefited from a long season. This immediately casts a shadow over any attempt to equate success with population growth at the seaside in any simplistic way: the dominant interests in these resorts had clearly chosen a different path and defined success in their own comfortable way. Nevertheless, the available evidence on visitor numbers is not robust enough for us to use it systematically to qualify the conclusions we draw from trends in resident populations, and it is now time to assess what the factors were that seem to have been most conducive to population growth in English and Welsh seaside resorts between 1911 and 1951.
The roots of success
As we have seen, the first general essential was an informal atmosphere and a welcoming attitude to the expanding popular market, as suggested by openness to plotland and holiday camp incursions. It is significant that, in general, resorts which had been dominated by large landed estates who had sought to safeguard a high 'social tone' did not expand impressively in this period: they found it difficult to move down-market while many of their traditional visitors were lured abroad (Cannadine, 1982). Accessibility continued to be crucial and this applied to retirement (which was becoming increasingly important) as well as to commuters and the expanding popular end of the holiday market. The motor-car was still used mainly for shorter journeys, and the train continued to dominate, although resorts like Blackpool which lay within easy reach of population centres saw rapid expansion in charabanc and car traffic as a supplement to the railways during the inter-war period (Brunner, 1945).

Seaside commuting was a widespread ingredient in resort growth in the inter-war years. The four top resorts which remained far ahead of the rest in population terms (Brighton, Southend, Blackpool and Bournemouth) all had large and growing commuter populations, as did the next three in the 1951 table (Southport, Hove and Worthing). All except Southport, which was fifteenth, were among the top six seaside resorts in the incremental growth table. Among other rapidly-growing resorts with commuters prominent in their residential profile were Morecambe, with its well-patronized 2-hour rail link with Bradford, Whitley Bay, which was a Newcastle dormitory, and Hoylake, serving a similar function for Liverpool and Birkenhead (Barke, 1992). Among the fast-rising resorts from humble origins, Canvey Island and Lancing developed dormitory functions at an early stage. An illustration of the difference commuters could make when most other things were equal comes from the Sussex coast and reinforces the theme of the importance of comparative accessibility. On the one hand, the number of season ticket holders between Brighton and Hove and London had increased to 2436 by 1938, boosted by the electrification of the route in 1933. Nearby Worthing, a similar beneficiary, had 497 such commuters. This was a dynamic part of the Sussex coastline in the inter-war years, although it could not match Southend's totals of 10,000 in 1929 and 10,500 in 1938. Further east in Sussex, Hastings stagnated, Eastbourne experienced only limited growth, and Bexhill failed to sustain its turn-of-the-century expansion. Significantly, Hastings was equal 156th and Eastbourne equal 152nd of the 157 large towns in Moser and Scott's survey in terms of the local importance of commuting in 1951. A slower rail service over a longer distance from London clearly played its part here (Brunner, 1945; Moser and Scott, 1961; Gilbert, 1954; Stafford and Yates, 1985).

Commuters were rarely the most important element in the burgeoning residential populations of many resorts, however. People of independent means who could afford to live for all or part of the year at the seaside were desirable inhabitants and potential big spenders in the local shops. Thus the new estates of substantial detached houses at Aldwick, on the western side of Bognor, which developed after King George V's convalescence in the area in 1929, were seen as a great asset locally, symbolized by the conspicuous forays of the society beauty Lady Diana Cooper to the shops in her 'big cream car' (Young, 1983). Access to London, in particular, was important to such wealthy residents as well as to the commuters, although Bognor Regis was unusual in sustaining such an up-market residential presence at a time when the Continent and especially the French Riviera beckoned.

Much more important to most resorts was the growth of seaside retirement, which increased steadily in significance and spread further down the social scale throughout the period. The proportion of people of pensionable age in the population of England and Wales more than doubled between 1901 and 1951, and a growing percentage of these had sufficient resources to contemplate retirement to the seaside, as occupational pensions became more widely available to top up state pensions and savings. Accessibility was important for keeping in touch with family and friends, but the attractiveness of resorts as retirement centres owed more to their own perceived qualities than to differences in the length of journey times. The Sussex experience certainly suggests this, for Bexhill's retired as a percentage of the town's population were not far behind Worthing's in 1951, when the latter was already the capital of Sussex's emergent 'Costa Geriatrica'. The four large towns in England with the lowest percentage of their over-15s in occupation were Worthing, Hastings, Hove and Bournemouth, with figures of between 43.6 and 49.4. Between 1901 and 1951 the gap between the proportion of people of pensionable age in the populations of seven Sussex resorts and the national figure widened steadily, accelerating between 1931 and 1951. By 1951 the proportion in Worthing stood at 29.7 per cent (compared with 9.2 in 1901 and 18.7 in 1931). The corresponding figures for Bexhill were 28.0, 6.3 and 15.0. Similar, if (usually) less spectacular, trends were noticeable on other coastlines, especially South Devon, where some of the smaller and more exclusive resorts, such as Budleigh Salterton, combined modest but sustained population growth with an increasing over-representation of the elderly. The retirement trend was evident in Cornwall and on other more peripheral coastlines, but had yet to make a substantial demographic impact, and it was on the coasts of Lancashire, North Wales and Essex that convenient access and familiarity boosted retirement from their populous hinterlands to resorts with limited claims to scenic and climatic distinction, although Morecambe did try to make its case as the 'Naples of the North'. By the second quarter of the twentieth century, anyway, seaside retirement was one of the most important components of resort growth in most English and Welsh regions, as the foundations were laid for even more far-reaching expansion in this sector in the 1960s and 1970s (Karn, 1977; Moser and Scott, 1961).

Resorts which attracted retirement populations included many small and (in population terms) stagnating centres, from Grange-over-Sands in the North-west to Southwold in East Anglia. It clearly shored up ailing resort economies as well as, in places like Worthing, boosting growth which was occurring for other reasons. To complicate matters further, places like Herne Bay (especially), Clacton and Morecambe, which were particularly lively during the inter-war years, were attracting extensive retirement populations by the 1950s, and no doubt these phenomena were connected. The rise of seaside retirement was part of a broader shift in many resort economies away from holiday industry functions as such, towards a less-specific residential and services base. This was to be the future for most seaside resorts, but by mid-century it had gone further on some coastlines and in some places than in others. Meanwhile, the holiday industry itself was capable of innovating, expanding and capturing new publics, and there is a good fit between high levels of population growth in individual resorts and the provision of holiday attractions in response to new demands.

We saw that resorts which could accommodate the new informality of plotland settlements and holiday camps showed particularly strong growth profiles over the period: indeed, the most dynamic of the small resorts, Canvey Island, was almost entirely composed of this kind of development. We can also, and perhaps relatedly, emphasize the importance of a combination of natural advantages, adaptability to new holiday styles, and readiness to invest in new amenities on a substantial scale, as attributes which accompanied above-average growth rates in resort populations.

A strong trend began the in Edwardian years, and continued during the inter-war years, towards more informal and less regulated use of the beach. Bathing machines and their associated regulations, with the attendant identification of sea-bathing as associated more with a prescribed health regime than with spontaneous and sustained pleasure, gave way to tents and an increasing tolerance of decorous undressing on the beach. Even more important was the rapid rise of the cult of sunshine and fresh air, which made the sun something to be courted rather than sheltered from, and made the possession of a sun-tan a social asset rather than a stigma. The strong positive association between sun, open air and health, which was gaining ground in the early twentieth century and ushered in the cult of sun-bathing after the First World War, had implications for holiday traffic flows. It increased the importance of a sandy beach with a regulatory regime which permitted the systematic exposure of hitherto hidden parts of the anatomy, and over long periods rather than fleetingly, reinforcing notions of the beach as a liminal zone where behaviour which elsewhere transgressed norms of approved conduct became acceptable and even expected (Stafford and Yates, 1985). In this context it is interesting to observe the large numbers of resorts with extensive sandy beaches and in the climatically-favoured South and East occupying prominent places in the percentage growth league table. Among these popular beach resorts, catering for the sort of lower middle-class and working-class families who could not dream of going abroad, were Skegness, Herne Bay, Walton-on-the-Naze, Mablethorpe and Withernsea. Newquay and Bude should also be included, and among resorts of this kind, growing rapidly from more established antecedents, were Bognor Regis, Clacton, Paignton, Weymouth and perhaps Whitley Bay. Margate should probably also be included, because its holiday seasons seem to have flourished in the inter-war years and its population growth figures may have been depressed by its particularly appalling experiences in the Second World War (Brunner, 1945). Any temptation to isolate climate and physical geography as determining factors brought into play by the new seaside culture should be resisted, however. Morecambe's success defies such expectations, while Prestatyn, Abergele and Rhyl showed that the key features could be reproduced in North Wales, and the big resorts had dynamics of their own by this time. They too had to keep up with changing times and meet new expectations, but places like Blackpool and Scarborough had the resources and municipal political cultures to make the necessary investments. Moreover, any emphasis on the South and East as favoured coastlines (before the Second World War turned them into front-line coasts and wrought severe damage to their economies) needs to allow for the better economic performance of these areas in terms of wages, employment and the capacity to generate comfortable retirement incomes at this time. In this context, however, the continuing buoyancy of the Lancashire coastal resorts merits further research. A pattern related to sunshine and beaches can be discerned, but there are many cross-currents.

The role of local government
The role of local government must also come into the equation. Local authorities were already adopting a particularly interventionist role at the seaside in the late nineteenth and early twentieth centuries, moving into recreational and entertainment provision as well as building and maintaining promenades and laying out parks and open spaces. Resort municipalities were more likely to operate utilities like gas, water and electricity works than their inland counterparts, and efforts were being made to move into advertising and publicity, although Blackpool remained unique in being allowed to levy (since 1879) a two-penny rate for advertising purposes. This was not the 'municipal socialism' beloved of the Webbs, although a lot of its outcomes were the same. It aimed at advancing the interests of the resort as a business collective, in competition with rivals at home and abroad, and needing public provision and operation of activities and amenities which would not have attracted private investment or for which private enterprise was seen as unsuitable in the context (Walton and Walvin, 1983). This was, in fact, what was later to be labelled 'municipal Conservatism', operated by overwhelmingly Conservative local government bodies and providing support services for the small businesses which dominated resort economies and electorates (Cooke, 1989).

The inter-war years saw the further development of these policies, especially in the largest resorts with most resources, but to some extent almost everywhere. This was particularly noticeable where landed estates ceased to exercise a positive and distinctive influence on resort development, which happened almost everywhere after the First World War, although (for example) at Folkestone Lord Radnor built an amusement centre next to the Marine Gardens to supplement the local authority's activities (Stafford and Yates, 1985). The 1920s and 1930s were probably the high point of local authority investment in seaside amenities: promenades, parks, pavilions, swimming pools, sun-terraces, sea defences and, in some places, grandiose redevelopment schemes. There was municipal involvement in entertainment and catering, and as late as 1937 Bognor Regis joined established up-market rivals in establishing a municipal orchestra (Young, 1983). There were golf courses and tennis courts, deck-chairs and bathing huts. It is clear that resort authorities took pains to provide the amenities which were identified with the new, open-air, athletic and sporting fashions in holidaymaking, as well as expanding their capacity to cope with expanding visitor numbers, a trend which seems to have been interrupted only briefly by trade depression in the early 1930s on some coastlines (Morgan, 1992). Meanwhile, the Health Resorts and Watering Places Act of 1921 allowed the expenditure of up to a penny rate on advertising resorts and their attractions, to be taken from the profits of municipally-run visitor amenities. This ushered in a period in which an increasingly professional attitude was taken to resort publicity, with the municipal holiday guide-book acting as showcase in an increasingly competitive world (Stafford and Yates, 1985).

These municipal initiatives did not always go unopposed, especially during the financially stringent early 1930s, and at times there was sustained ratepayer and residential opposition to perceived high levels of spending, and to alleged mismanagement of municipal catering and entertainments (Morgan, 1992). Seaside resorts took eight of the top 11 places in comparative per capita expenditure on parks and open spaces in English and Welsh county boroughs between 1919 and 1939, with between 8 1/2 times (Blackpool) and just over twice the national average. The three remaining places were taken by Wallasey, which included New Brighton, and West Hartlepool and South Shields, which contained resort areas within their boundaries. The three biggest per capita spenders on highways, which included promenades and sea defences, were Blackpool, Brighton and Hastings; but, significantly, seaside resorts were a long way down the municipal housing expenditure table. Business priorities ruled: considerations of social welfare were sidelined, or left to be solved by the provision of employment through economic growth achieved by other policies. This was the other side of the logic of 'municipal capitalism' (Ward, 1988; Morgan, 1992).

The importance of local government in resorts is undeniable, and a high level of investment in the holiday industry and related amenities was necessary everywhere to sustain competitiveness. On the evidence so far available, however, there seems to be no clear relationship between municipal enterprise and success as measured by population growth. Both Morecambe and Bognor Regis had active local government bodies, but criticism has been directed at them for (in Morecambe's case) parsimony towards popular festivals and a preference for showy municipal premises over holiday industry investment' and (in Bognor) failure to obtain and control the development of key sites or to provide a full range of municipal attractions (Simmill, 1993; Young, 1983). Scarborough and Hastings, on the other hand, were at the forefront of innovation across the board, but did not see their endeavours reflected in population growth.

This emphasis on municipal activity should not obscure the continuing importance of private enterprise in resort development. This was most generally obvious in housing and hotel provision: seaside resorts were generally slow and grudging in providing council houses, although Brighton laid out extensive estates to the east of the town for relocated slum dwellers, building 2700 houses between 1920 and 1939. This was a unique scale of response to a uniquely obvious set of problems (Gilbert, 1954). New houses were provided overwhelmingly by private developers everywhere, while new hotels brought highlights of modernity to inter-war sea fronts, the Midland Hotel at Morecambe being particularly outstanding, while hotel-building activity was especially in evidence at Torquay (Morgan, 1992). There was also continuing investment in large-scale entertainment complexes, with Blackpool's Pleasure Beach (brought up to date with new streamlined styles in the late 1930s) and Margate's Dreamland well to the fore (Turner and Palmer, 1976; Stafford and Yates, 1985). Nor should we ignore the lesser entrepreneurs at the bottom end of the market, such as Blackpool's Luke Gannon, the impresario of freak stalls and starving honeymoon couples on the Golden Mile, whose heyday in this respect came in the 1930s (Blackpool Gazette, 1935).

Private investment in entertainment and amenities may have been drawn disproportionately to the fastest-growing resorts, although more research would be needed to test this and to enquire into cause and effect, but the inconclusive nature of the findings about the relationship between municipal enterprise and population growth suggests that it is time to look more critically at the definitions of success which have so far set the tone for discussion. As the evidence on visitor numbers suggested, there is more to success than population growth, especially when we focus in on the holiday industry itself rather than the towns as a whole.

Assessing outcomes
An alternative measure of success might emphasize the achievement of steady prosperity or comfort without encountering the problems associated with growth and mixed visiting publics which the larger resorts encountered. On this measure, for example, the socially exclusive resorts of South Devon might be seen as highly successful on their own terms. It would be hard to argue that Salcombe (for example) was a failure on these criteria: it was highly successful in its own niche. Eastbourne is another interesting example. It grew only slowly in population terms, but it kept its architectural amenities and gardens, and in 1950 it had the fifth highest figure among Moser and Scott's 157 large towns for retail expenditure per head, behind Worcester, Bournemouth, Oxford and Bedford (Moser and Scott, 1961). Torquay poses the interesting problem of a place which grew quite impressively during the period but Went down-market in the process, while alongside its new hotels there was a rapid expansion of the boarding-house sector, and by the late 1930s half the town's houses provided some kind of accommodation for the August visitor peak (Morgan, 1992). For whom is this success, and on what terms? All this illustrates the dangers of concentrating on size and development in terms of the 'resort cycle' at the expense of other relevant concerns (Goodall, 1992; Walton, 1994).

In some ways the seaside resort remained a generically distinctive kind of town at mid-century, despite the varying fortunes to which this chapter has drawn attention. Moser and Scott (1961) pick out some of the salient features among the larger resorts. Seaside resorts were likely to have high percentages of the over-65s in their populations, and low percentages of the under-15s. They had considerable surpluses of women over men, especially the retirement resorts of the Sussex coast. They had low birth-rates, but some had among the highest illegitimacy rates, suggesting perhaps the attractions of anonymity as well as the moral perils of the season and the transient labour market. They had low levels of overcrowding (at least when the census was taken) and high concentrations in professional and service occupations. They were also, as Gilbert (1954) noted, particularly likely to elect Conservative MPs by large majorities and on the basis of low turnouts. We should also emphasize that resorts had their share of poverty and seasonal unemployment. Robert Noonan's, or Tressell's (1910), account of the building trades in Edwardian Hastings, or George Meek's insecure and poverty-stricken career as the only socialist bath-chairman in Folkestone at the same time, are borne out as enduring experiences by subsequent evidence on seasonal unemployment and poverty in the larger resorts. Seaside resorts were difficult places for labour to organize, and neither trade unions nor the politics of the Left flourished, although there were brief stirrings in the aftermath of the First World War. The defence of order and protection of an image of consensual respectability were prime concerns of resorts' rulers, although gang activity had a high profile at times in inter-war Brighton (Walton, 1992; Gilbert, 1954). All these might be regarded as generic qualities of seaside resorts which helped to give them a collective social identity.

On the other hand, however, attention should be paid to Morgan's (1992) suggestion that the years after the First World War saw resorts in Devon, and perhaps elsewhere, diversifying to cater for different markets. This was already a feature of coastal development over much of England in late Victorian times, in fact, and there is no doubt that within a framework of shared experiences and identities the seaside resorts of England and Wales continued to offer a remarkable diversity of holiday and residential environments. This diversity, which may indeed have become more marked during the period as new kinds of provision emerged, probably played a central role in the continuing growth and proliferation of resorts in the first half of the twentieth century. In spite of foreign and domestic competition, the English seaside resort remained remarkably dynamic during this period, and enough towns managed to combine traditional attributes with openness to innovation to ensure that most established resorts did more than survive: they prospered, although this prosperity did not extend to all their inhabitants or workers. The traditional boarding house seaside holiday, based on a railway journey to an individual resort, still held sway in 1950 as it had in 1900. What the future held was a different matter.