1 Introduction

The aim of this paper is to examine the relationship between primary and second homes within the overall context of the Spanish housing market. In the academic literature on second homes such dwellings have conventionally been treated as a separate sub-segment of the broader housing market. The second home has therefore been seen as distinct from the primary home. The justification for this distinction usually relates to the identification of the differentiating characteristics of second homes. For example, their locational features (Rogers 1977; Hoggart and Buller 1995; Hernández Borge 2007), the socio-economic profiles of owners (López Colás and Módenes Cabrerizo 2005), their role in seasonal and longer-term international migration in particular parts of the world (Smith and Duffy 2003; Hall and Müller 2004; Díaz Orueta and Lourés 2006), and the part they may play in local areas through changing the structure of communities and conflicts over housing resources (Downing and Dower 1973; Coppock 1977; Gallent and Tewdwr-Jones 2000). These, and other ‘internal’ characteristics of the second home market have dominated the literature. Much rarer is the consideration of links between second homes and wider housing market issues with, for example, the study of the demand for the limited amount of dwellings in certain areas and the subsequent ramifications for the rest of the housing market (Shucksmith 1983; Paris 2006) being one of the comparatively few acknowledgements of this broader housing market context. Few studies have sought to systematically explore the linkages that may exist between first and second homes and it is this exploration that forms the primary aim of this paper. It will be argued that, in the context of Spain at least, second homes are an integral part of the overall housing market.

The examination of the relationship between first and second homes is given particular significance through the evident lack of consensus within the limited literature on the nature of this relationship. On the one hand, authorities such as Ball (2005, p. 123) have argued specifically in the case of Spain that “The secondary home market operates independently of the primary one…”. On the other hand, Manrique and Ojah (2003) assert that “… demand for both types of housing are, indeed, related and they are both considered necessities; thus suggesting that their consumption ought to be modelled as a joint decision” (p. 323). Similarly, López Colás et al. (2007) argue that the purchase of a second home is frequently a natural continuation of a family investment strategy, initiated with the purchase of the primary home. The inter-relatedness of primary and secondary homes within the construction sector is also inferred by Cabrera (2004) who notes that “All experts agree that construction emerged in the nineties as the main driving motor of the Spanish economy, together with tourism, and in large measure connected to the latter, as a good part of housing is purchased as second residence.” Thus, we have two apparently contradictory statements on the nature of the relationship between the market for primary and second homes in Spain, a situation that adds particular resonance to the aim of subjecting the precise nature of the relationship to closer examination.

However, several more detailed research questions arise in pursuance of this aim as we need to explore the precise nature of some of the relationships. First, it is widely recognized that, over the medium term period of more than a few years, the housing market overall is highly cyclical with periods of expansion alternating with periods of relative contraction (Ball and Morrison 2000; Taltavull de La Paz 2000, 2003). Therefore, a question that arises is—does the market for second homes behave in the same way as that for primary homes in relation to economic cycles of boom and slump? Second, given that second homes are widely held to be a response to the demands of a specific market niche, especially related to leisure and vocational pursuits, it is legitimate to ask whether or not the market for second homes is differentiated by settlement type and type of dwelling. Third, do the specific features traditionally identified as characteristic of second homes lead to a second home market that is spatially differentiated from that for primary homes? Fourth, given the significance of international tourism in Spain’s economy (Tamames 1986; Salmon 1995), does foreign ownership of second homes introduce a significant differentiating factor that ‘distances’ processes operating in the second home market from those in the primary home market? Finally, in the knowledge that Spain has undergone quite remarkable social and economic change over the last three or four decades (Giner 1972; Shubert 1990; Gonzalez and Requena 2005), what are the implications of this rapid social and economic transformation for the relationship between second and primary homes? In particular, does this have fundamental implications for the potential breakdown of ‘traditional’ attitudes towards land and property (Kenny 1966; Brandes 1975; Behar 1984; López Colás et al. 2007) and their replacement by more universal contemporary norms?

2 Data problems

It must be acknowledged that attempts to assess the relationship between primary and second homes in any housing market are bedevilled by the lack of precise and suitable data. Recourse has frequently to be made to data sources that provide useful but often indirect and inferential insights into relationships. Furthermore, any investigation is also hampered by the ambiguity and different usages of the term ‘housing market’. The term is a frequently used one, both in everyday and academic discourse, but is one that encompasses several components. For example, reference to the dynamics of the housing market in any particular territory could potentially relate to empirical evidence on new build, on housing transactions, or on actual house and rental prices. All three may be construed to provide evidence on the nature of supply and demand but the three components are, in several ways, differentiated from each other and measure related, but different, aspects of the housing market in any territory. Ideally, any empirical examination of the relationship between primary and second homes would analyse each of these three components. Unfortunately, data is not readily available for primary and secondary homes for all three of the components of the housing market identified—construction, transactions and price levels—and the paper will concentrate on the data available from official Spanish statistical sources. New build data is available and, although not differentiated between first and second homes in the sense that it is not possible to say how many new build homes are for primary or secondary use, the Spanish housing census does combine date of build with status (i.e. primary, second home or empty) in the census year. In other words, information is available on the year of construction of second homes but it is not necessarily the case that the dwelling was purpose-built as a second home in that year. Data on the number of new mortgages taken out are available in Spain but data on the number of housing transactions are not readily accessible and certainly not for second homes differentiated from primary homes. House price data are also available but with no differentiation between primary and second homes.

Therefore, empirical data on second homes is somewhat problematic (Casado-Diaz 2004) but the most comprehensive source is the 2001 census of population and housing in Spain. These data do present a number of problems, however, of which the most significant is probably the potential for confusion between ‘empty’ and ‘secondary’ homes (Miranda Montero 1987; Barke 1991). It has been suggested that, in both the censuses of 1981 and 1991, dwellings that were, in fact, secondary homes were counted as empty by census enumerators (Valenzuela 2003a). Within the Spanish census, there are two sets of data relating to second homes. The first is based on what, in fact, is virtually a separate but contemporaneous census, enumerating buildings rather than population and classifying as secondary family dwellings, housing that is not considered to be the principal residence and which is used for only part of the year. The second source of data is derived from the questions asked of households where, crucially, the respondent is asked about the availability and use, not ownership, of a second residence and where that residence is located. The information on the former is therefore clearly based on the dwelling as a physical structure; in the latter case it is based on households and their access to, rather than ownership of, a second home.

Before proceeding to an empirical examination of what these data can tell us in relation to the research questions posed earlier, however, it is necessary to interrogate the limited literature on the theoretical relationships between primary and secondary housing markets.

3 Theoretical perspectives

As noted earlier, much of the literature on second homes has tended to assume that they constitute a relatively discrete component of the overall housing market and therefore it is not necessary to discuss their relationship to that wider housing market. First and second homes have tended to be seen as separate entities. The main reason for this is that, until relatively recently, “… most scholarly writing on second homes derives from leisure and tourism studies, rural studies, planning and cultural studies … the ‘housing’ literature contains little about second homes” (Paris 2006, p. 3). This relative neglect has served to reinforce the perception of second homes as a ‘luxury’ commodity, a perception that, arguably, has limited the interpretation of second homes in the context of the overall housing market. For example, one of the few theoretical perspectives on the possible relationships between primary and secondary homes is Ball’s account of the Spanish housing market (2005) which claims a considerable degree of independence of the two. However, the initial basis for this assertion is rather unclear. Immediate reference is made to the secondary home market being “… influenced by domestic economic conditions and the economic situation in countries such as Germany, the Netherlands and the UK, because many purchasers are foreign nationals” (ibid). Clearly, the primary home market is also surely affected by domestic economic conditions, leaving the burden of the explanation of the difference between the two components with the significance of foreign purchasers. Elsewhere, however, (p. 14) Ball argues that the market for second homes is different from that for primary homes in the important sense that it is considerably more volatile. Four principal reasons are cited for this difference (Ball 2005).

3.1 Demand is more discretionary than in primary home markets

Second homes are not as essential so when times are hard they are more likely to be sold first or vendors are more willing to accept lower prices.

In general terms this assertion has a logical validity. However, there may be circumstances where ‘hard times’ may demonstrate the strength of the link between first and second homes. For example, much may depend on stage in the life cycle. It is quite feasible that, when ‘times are hard’ some, especially those on the verge of retirement or recently retired, may choose to sell their primary home and move to their second home (Williams et al. 2004). In such cases there is a clear housing market link between the two related to life-course planning and household investment strategies (Hall and Müller 2004; Smith 2005). Others, regardless of any impending retirement, may choose to ‘downsize’ from a primary home to the former second home. Recent research in the USA has demonstrated that a substantial proportion (28%) of vacation home buyers intended that this would become their primary home in the future (National Association of Realtors 2007). Much is also likely to depend on the relative locations of the workplace and primary residence. It is quite possible that the second home may actually be more conveniently located for work than the primary home. An apartment in the city, purchased as a second home/urban investment with the main family home being more distantly located in relation to workplace is a phenomenon that is much more common in Spain than in northern Europe (Valenzuela 2003b), hence the considerable number of urban second homes. But the latter may be retained in preference if ‘downsizing’, or an adjustment of household investment strategies, become necessary. Research in Spain has shown that, contrary to northern European behavioural norms, in ‘high price areas’, some households have a greater tendency to rent their primary home and own their secondary home (Manrique and Ojah 2003). The main point, however, is that second homes, rather than being a ‘luxury’ item that may be readily discarded, are frequently an integral part of a wider strategy of family related investment.

3.2 Lenders are more likely to be concerned about defaults

It is difficult to see the basis for this assertion as no empirical evidence is produced to suggest that lenders are ‘tighter’ with second home loans than with loans for primary residences. Competition between and deregulation of financial services agencies has led to credit being readily available for all sorts of purchases in recent decades, including second homes (Smith 2005). “In the case of Spain, over the past few years, debt taken on by households to acquire housing has risen to a level equivalent to 60% of gross disposable income, a figure that has doubled in the last 10 years” (BBVA 2005, pp. 21–22). There is no evidence from Spain that lenders are more reluctant to make loans on second homes than they are for the purchase of primary dwellings.

3.3 There is no fallback demand

In the primary market some households are priced out during booms, especially first time buyers. In downturns, affordability improves and first time buyers eventually may be drawn into purchase, thus ameliorating potential falls in demand. Apart from a limited number of ‘bargain hunters’—in periods of relative depression—second home owners have no such group to fall back on.

The suggestion here is that, in downturns, first time buyers are more likely purchasers as affordability has improved for them. Yet, if affordability improves for first time buyers, it may equally improve for others, for example, those intending to sell an existing house and move on elsewhere. But much is likely to depend on the nature of the housing stock available in particular localities as first time buyers tend to be attracted to particular types of dwellings and locations. Therefore, the argument that first time buyers represent a significant ‘fallback’ market for primary homes is not very persuasive and thus the case that first and second homes are different in this respect is not sustained. But more importantly, ‘affordability’ for new buyers (the ‘fallback demand’) will only really take place if there is a significant decrease in price, that is, a price crash. Such price crashes are relatively rare events and ‘soft landings’ are much more normal, that is, a levelling off of the rate of increase. In which case, ‘affordability’ is not likely to dramatically improve for first time buyers.

3.4 Supply-side is more likely to transmit volatility

‘Feast or famine’ is more likely in second home markets than primary ones. When supply is low, scarcity will lead to rapid price increases but when supply is plentiful, developers may overestimate demand and eventually be forced to sell off at low prices. Ball suggests that the second home market has more price volatility than markets with more normal supply environments.

The assumption underpinning this point is that developers are more responsive to the situation in primary markets than they are to that in the second home component. However, such a scenario would only occur if there was a significant difference in the market intelligence relating to primary homes and secondary homes but there is no immediately obvious reason why this should be the case. In both cases the majority of building is speculative and it seems equally likely that over- or under-estimates of demand could occur. Again, therefore, it is difficult to see the real basis for this differentiation of primary and secondary housing markets.

These four factors represent arguments for the potentially different behaviour of the secondary home market when compared to that for primary dwellings but their basis is clearly problematic and falls short of a rationale for the complete autonomy of the two sectors. Indeed, contemporary research (McIntyre et al. 2006) has begun to explore the inter-connections between first and second homes, stressing the need to see them as ‘linked spaces’. Structural changes in late modern society have important implications for the primary-secondary home distinction with the boundary between the two arguably becoming increasingly blurred. Perkins and Thornes (2006) argue that “Primary and secondary homes are not therefore polar opposites, but rather they represent a continuum” (p. 79). Thus, the traditional hierarchical distinction between ‘first’ and ‘second’ home may be less relevant in contemporary society and increasingly be replaced by the concept of multiple homes and, significantly, in the case of Spain, Manrique and Ojah (2003) have claimed that “… Spanish households consider both primary and secondary homes as necessities” (p. 335).

It is against this contested theoretical background that this paper seeks to explore the various ways in which primary and secondary homes in Spain are linked, specifically in relation to the five research questions identified earlier, namely:

  • Does the market for second homes behave in a similar way to that for primary homes in relation to economic cycles?

  • Is the market for second homes differentiated by settlement size and dwelling structure?

  • Is the market for second homes spatially differentiated from the market for primary homes?

  • Does foreign ownership of second homes differentiate this component of the housing market from the primary housing sector?

  • What are the implications of rapid social and economic transformation for the relationship between second and primary homes?

4 Does the market for second homes behave in a similar way to that for primary homes in relation to economic cycles?

Before attempting to analyse this issue in detail it is necessary to establish the scale and development trajectory of second home ownership in Spain. Over the last half century the total number of dwellings in Spain increased by 226% but the number of second homes increased by over seven times that percentage. Robinson (1990) asserts that, since 1945, Spain has experienced the fastest growth of second homes in Europe. The 1970s saw a particularly high numerical increase in second homes, although since then, growth has continued with annual increments in excess of 70,000. Due to the large number of second homes in Spain, the proportion of housing that exists in the country is well above the European average, 510 houses for every 1000 inhabitants compared to 445 per 1000 (BBVA 2003). Furthermore, about one million housing units are sold annually in Spain, an estimated two-thirds for private use and one-third for investment, the latter being predominantly second homes (El Pais, 5th May, 2003). Up until the recent past, no differentiation was made between first and second homes in the availability of income tax relief on the cost of purchase (Pareja Eastaway and San Martin 1999; Lopez-Garcia 2004; Cabré and Módenes 2004), in effect blurring the differences between the two and, for many indigenous Spaniards, possibly distancing the ownership of a second home from its interpretation as a ‘luxury’ commodity. In the second half of the twentieth century, therefore, second homes have clearly become a highly significant component of the overall Spanish housing market (Table 1).

Table 1 Total dwellings and second homes in Spain, 1950–2001 (exclude Ceuta and Melilla)
Full size table

The 2001 census enumerated 21 million dwellings in Spain with 16% being recorded as second homes. Not surprisingly, the annual rate of increase for three decades after the mid-twentieth century considerably exceeded the rate of increase of primary dwellings, suggesting that, for this period at least, these two components of the housing market were, in aggregate, behaving in rather different ways. Yet, for individuals, there may have been very logical and deliberate links between first and second homes. The “intense migration from rural to urban areas in the 1960s and 1970s led to a large increase in empty houses, from 4.7% of the housing stock in 1960 to 16.3% in 1981 …” (BBVA 2003, p. 7). A significant proportion of these empty houses were subsequently destined to become second homes as ownership was retained and improving standards of living made it possible to use such dwellings for vacations and week-ends (Kenny 1966; Brandes 1975; Aranda Palmero 2002). The 1980s saw the highest growth of second homes in relation to primary dwellings, mainly because the rate of increase in the latter declined substantially. However, the final decade of the twentieth century saw a convergence in the rate of increase of both sectors of the housing market, possibly through the market correcting the over-supply of second homes in the previous decade.

A useful, if indirect, measure of the relative volatility of the primary and secondary housing market is indicated by rates of new build. Figure 1 compares the growth trajectories for newly built primary and secondary homes in Spain through the period 1981–2001. In broad terms there does appear to be some relationship between the two with peaks being apparent in the mid 1980s, 1989/1990 and 2001 and troughs in 1983/1984, 1986/1987 and 1995. However, the statistical correlation between the two is barely significant at R = 0.339 and Fig. 1 indicates a marked difference between the 1980s and 1990s with the relative significance of second homes being much greater in the former decade than in the latter. The considerable slump in the 1990/1991 period was particularly marked in the case of second homes. The early 1990s was a period of house price slump generally (Pagés and Maza 2003) in Spain but there is a strong suggestion here of an excess of supply of second homes building up in the later 1980s leading to a surplus by the end of the decade and consequently to a rather more cautious attitude through the 1990s. Although the significant slump in the 1990/1991 period does point to a greater degree of volatility in the second home market, the relationship between the two is otherwise a reasonably close one. For housing markets generally in Europe, the early 1990s were a rather unusual and not particularly typical period.

Fig. 1
figure 1

Primary and second homes building index, 1981–2001 (1981 = 100)

Full size image

But data on new building represents only one dimension of the supply side and the relationship between primary and secondary dwellings in the housing market. Another dimension relates to price levels and recent data suggest that “Indications of moderation are rising not only in the primary housing segment but also in the market for second homes” (BBVA 2006, p. 15), again suggesting a close relationship between the two.

5 Is the market for second homes differentiated by settlement size and dwelling structure?

Our second research question concerns the extent to which second homes constitute a discrete sector of the market in terms of type of settlement and type of building. Table 2 shows the distribution of second homes by size of municipality. It should be noted that ‘municipality’ in this context means a spatial unit of local government and is not synonymous with settlement. One municipality could contain several settlements of varying sizes.

Table 2 Total dwellings and second homes by size of municipality, Spain 2001 (exclude Ceuta and Melilla)
Full size table

Table 2 indicates that second homes are distributed across the range of municipality sizes in Spain with more than one quarter being found in municipalities of over 50,000 population. It is clear, therefore, that second homes are not just a phenomenon of rural areas. However, their relative significance is clearly much greater in smaller areas, as the final column of Table 2 shows. This would appear to support the notion of a significant proportion of second homes being a residual phenomenon, resulting from the process of rural depopulation. Further data on the age of second homes (Table 3) does not support this view entirely, as the ‘rural depopulation’ hypothesis would lead us to expect a very high proportion of second homes being of considerable age. At the national scale, this is clearly not the case as 60% of all second homes have been built since the early 1970s.

Table 3 Period of building of total dwellings and second homes in Spain, 2001 (exclude Ceuta and Melilla)
Full size table

The data above concerns only the period of building of the properties and it does not necessarily follow that buildings built within a certain period were purpose-built second homes. Clearly, many buildings built as principal dwellings could be converted into a second home at any time. Nevertheless, Table 3 suggests some interesting features. The age structure of second homes broadly parallels that of all dwellings, a feature which is suggestive of the pervasive character of the second home phenomenon in Spain and which adds at least some circumstantial support for the thesis that primary home ownership and secondary home ownership form a closely related phenomenon of the housing market (Manrique and Ojah 2003). The proportion of any age cohort of dwellings that are used as second homes is remarkably consistent and, at the aggregate level at least, serves to dispel any notion of second homes consisting predominantly of the ancient, dispersed former agricultural cottage or farm building nestling deeply in the countryside. However, in many rural locations old family properties have been demolished and replaced with new dwellings and, elsewhere, land that has been retained for several generations has been used to build a new second home. Whilst a near majority (43%) of second homes are detached, single dwellings (Table 4) almost as many occur in blocks of over 10 dwellings. There is also a higher proportion of second homes located in blocks of over 40 dwellings compared to the proportion of all dwellings in this largest size group.

Table 4 Number of dwellings in the building, Spain 2001 (exclude Ceuta and Melilla)
Full size table

Table 4 also confirms that second homes in Spain are not confined to just one type of physical structure. Just over one fifth of all single detached dwellings are second homes, but so too is the same proportion of dwellings in blocks of over 40 dwellings.

6 Is the market for second homes spatially differentiated from the market for primary homes?

Although much is made of the role of foreign tourism in the development of second homes (Casado-Diaz 1999; Díaz Orueta and Lourés 2006) and the role of foreign investment in real estate (Carrascosa and Sastre 1992), it is acknowledged that “In Spain … it is not the coastal areas that have the greatest shares of second homes but, rather, the provinces around Madrid. The highest at 42%, is Avila, followed by 38% in Guadalajara …” (Ball op. cit., p. 10).

Figure 2 shows the percentage of the total dwellings built in each province in the 1981–2001 period that were second homes in 2001 and therefore gives some indication of the active development and provision of second homes in relation to the market for primary dwellings. In some locations it has been argued that the residential tourism model, of which second homes and holiday homes are the most significant component, has features that are closer to the overall property development industry than to the tourism sector (Mazón 2006). But this ‘active’ development is far from being confined to the tourism-oriented Mediterranean coastal provinces. It is of interest that, of the 13 provinces with a significantly high proportion of second homes in 1981 (in excess of 0.5 of the standard deviation above the mean proportion), 12 had significantly high levels of ‘new build’ second homes in the subsequent 20 years. This ‘active’ development of second homes is widespread in a number of regions and, in the interior of Spain, many are not just residual features, consequent upon factors such as rural depopulation. In a number of predominantly rural provinces, for example, in Toledo, Huesca, Cantabria, Pontevedra and Lugo, over 40% of the second homes were built between 1981 and 2001.

Fig. 2
figure 2

Per cent of all houses built, 1981–2001 that were second homes in 2001

Full size image

The existence of a spatial dimension to housing construction cycles generally in Spain has been recognized previously (Taltavull de La Paz 2000) but an interesting issue is the extent to which different regions may exhibit varying relationships between first and second home construction. For each province we can compare the amount of new building for second homes in the 1981–2001 period with new build for primary dwellings. These relationships are best expressed as a ratio, in this case the numbers of second homes and primary dwellings built between 1981 and 2001 per 1,000 resident population in 2001. One of the recurrent arguments supporting the existence of a discrete market for second homes is their alleged concentration in predominantly coastal, tourist-oriented zones. This argument would be strongly supported if such provinces demonstrated a high ratio of newly built second homes per 1,000 resident population and a modest or low ratio of new build primary homes. In furtherance of the separate housing markets argument we may expect that, elsewhere, high ratios of primary house building would be accompanied by low or modest ratios of second home building. In fact, there is a positive relationship between the two ratios as indicated by a Pearson Product Moment correlation coefficient of r = +0.519, indicating that primary and secondary home new build is certainly not mutually exclusive, at least at the provincial level.

Spain’s 50 provinces were classified, using the standard deviation around the mean for the two variables, that is, new built primary and second homes per 1,000 resident population in 2001. Each province could then be categorised into high (in excess of one standard deviation above the mean), medium (within one standard deviation of the mean) and low (below one standard deviation from the mean) for both the ratio of new build second homes per 1,000 resident population and the ratio of new build primary homes per 1,000 resident population. No provinces exhibited polarised housing markets, that is, there were no provinces that were significantly high on one dimension and low on the other.

However, there is considerable variety across Spain in the relationship between new build second and primary homes. Díaz Orueta and Lourés (2006) stress the dangers of extrapolating conclusions reached in specific areas of the country to the whole of Spain, noting the high degree of autonomy, and the varied outcomes, in housing and planning matters now enjoyed by Spain’s Autonomous Communities. For example, in the category where both forms of housing development were significantly above the mean were two very different types of province, two Mediterranean coastal tourist-oriented areas, Alicante and Málaga and three interior areas that are well within Madrid’s long distance commuter range—Avila, Guadalajara and Toledo. Relatively low rates of new build of both components were found in some peripheral rural provinces, in the Basque country, but also in the cities of Madrid and Barcelona. Provinces with very different characteristics and locations share similar trajectories of development over the two decades indicating that the relationship between the construction of first and second homes is clearly not uniform over space and the nature of the relationship between primary and secondary homes in the housing market may be subtly different in different parts of Spain. What is clear, though, is that these data at the provincial scale do not suggest independent operation of the two components in a spatial sense at least. There are few provinces where the development of one component has been significantly high at the expense of the other component.

7 Does foreign ownership of second homes differentiate this component of the housing market from the primary housing sector?

One of the main differences between the primary and secondary markets suggested by Ball is that foreign investment and ownership constitutes a significant proportion of the latter. The key point in this context, however, is that there is also considerable foreign investment in the primary housing market. In the latter part of the twentieth century foreign direct investment has played an outstanding role as a driving force of overall economic development in Spain (Bajo-Rubio and Lopez-Pueyo 1997) with investment in real estate in particular providing an increasingly significant component (Carrascosa and Sastre 1991, 1992; Keogh 1996; Sastre 2002; BBVA 2003, 2005, 2006). Even before Spain’s full integration into the EU, such investment was significant with, for example, 10.4% of total investment in real estate in Spain between 1962 and 1990 being from foreign sources (Carrascosa and Sastre 1991, p. 144). Not surprisingly, the majority of this foreign investment has been concentrated in the coastal regions of Spain. For example, in 1988–1990 about 75% of investment in real estate was in the tourism-oriented regions of Andalucía, the Canaries, the Baleares and Valencia (Carrascosa and Sastre 1991, p. 140). However, such investment is far from being limited to second homes alone. Díaz Orueta and Lourés (2006) argue strongly that, whilst it is not possible to ascertain how many housing units remain as holiday homes or are sold as permanent homes, the increase in registered foreign residents in an area is a clear indication of the latter. Registration bestows many benefits for long term residents and an increase in registered foreign residents in an area is therefore a useful indicator of permanent residence. At the provincial level there are strong statistical correlations between total new house building over the 2001–2005 period and the absolute numbers of foreign residents in 2007 (r = +0.872) and also between total new house building and change in the numbers of foreign residents 2001–2007 (r = +0.866). Total house building bears a strong relationship to the distribution of foreign residents but this is house building for all purposes, not just for second homes.

Furthermore, if the role of foreigners in driving demand for second homes is so strong, we would expect clear relationships between the proportion of foreign residents and the proportion of second homes and the increase in foreign residents and the increase in second homes. Table 5 shows that the correlations of these variables for Spain’s 50 provinces are extremely low and fail to achieve any statistical significance. The same is true for the relationship between increase in foreign residents and the increase in primary residences. These data do not, of course, take account of the many foreigners who own second homes but who do not register as residents but the available empirical evidence suggests (Rodríguez et al. 1998; King et al. 2000; O’Reilly 2000; Salvá Tomás 2002; Casado Diaz 2004; Díaz Orueta and Lourés 2006) that there is strong coincidence between the location of foreign owned second homes and the location of foreign owned primary homes, purchase of the former frequently leading to conversion to the latter. “The trend is for periods of occupancy in these second residences to grow longer, and often these second residences turn into primary homes” (Live in Spain 2006). In this respect, at least, the two markets are similar.

Table 5 Correlations between foreign residents and secondary and primary homes, 1981–2001, Spain’s provinces
Full size table

There are, of course, good reasons why this should be the case. Foreign demand is—like domestic demand—not limited to the acquisition of second homes alone but consists of purchasers who come to live permanently, and who buy as an investment with an eye to future profit, as well as purchasing for seasonal use (Díaz Orueta and Lourés 2006). Furthermore, the once clear distinctions between these reasons for purchase have become blurred which is another argument against the existence of a ‘foreigner—driven’ market for second homes that is totally separate from that for primary dwellings.

It is demonstrably the case, however, that there is a spatially differentiated foreign owned component in the Spanish housing market and the fact that this is so large has important implications for the Spanish housing market in aggregate. The strongest dichotomy in the contemporary Spanish housing market may be between Spaniards and non-Spaniards rather than between primary and secondary homes, and this dichotomy is particularly manifest in a spatial sense with the concentration of foreign residents in both forms of housing on the islands and ‘costas’.

8 What are the implications of rapid social and economic transformation for the relationship between second and primary homes?

The ‘modernisation’ of Spain in the latter part of the twentieth century is a recurrent feature of scholarly works on that country’s social and economic history. But, to what extent are such changes manifest within recent developments in the Spanish housing market? The spatial pattern of second homes in Spain is a ‘mosaic’ consisting of different layers of development and change, in some areas the pattern representing traditional processes whilst in others significant contemporary social and economic change has had a major impact (Barke 2007). One indication of these different processes concerns the issue of how pervasive second home ownership is across the socio-economic spectrum in Spain. Educational levels may be used to define a broad status hierarchy (Hoggart 1997). Three grades can be recognized; education to primary level only, completion to second grade (usually age 18), and tertiary education including advanced professional training, diploma and degree courses and doctorates. Not surprisingly, the highest rate of access to a second home according to the 2001 census was in households headed by people with tertiary qualifications (28.3%). Just 17.1% of those with secondary education only had access to a second home and 13.5% of those with primary education or less did so. However, in most of northern Europe and North America it would be considered remarkable that over 13% of household heads with minimal or no formal education should own a dwelling apart from their primary residence. More significantly, this suggests a housing market with more limited internal cleavages than is the case in, say, the UK. “… Spain, as a society, has a widespread trend towards second home demand, independently of the socio-economic, regional or residential status” (Módenes and López Colás 2005, p. 10). Second homes in Spain are less likely to be considered a luxury item and are more likely to be fully integrated into the overall housing market.

In contemporary Spain, ‘new’ processes have been operating in addition to the ‘old’, the latter being particularly related to a historical tendency to ‘hang on’ to the old family home in the countryside, dating from the intensive rural-urban migration of the 1960s onwards (Kenny 1966; Brandes 1975; Elbersen 2005). Ros Tonda (2003) has drawn attention to national differences in demand for second homes, stressing the historico-cultural and life style aspects of Spanish second home ownership. It has been argued by others that “In Spain, the family network is crucial to the transition to ownership. One of ten owner-occupied homes is a legacy or a gift from family” (Cabré and Módenes 2004, p. 19). This applies equally to second homes as to primary dwellings. Keogh (1996) has observed that Spain possesses “… a distinctive cultural attitude to the long-term ownership and the intergenerational transfer of property and other assets.” An issue of importance in this context relates to the perception of second homes as an item of consumption. The ‘luxury item’ status of second homes is often implicit in many analyses of North American and north west Europe (Hall and Müller 2004) but other cultures may have a different perspective. López Colás et al. (2007) argue that because the family in Spain still retains considerable significance in the overall structure of social relationships, the inter-generational transfer of assets, including second homes, plays an important role in cementing such relationships.

If much of the data available from the Spanish census points in the direction of the emergence of a close relationship between the primary and secondary housing markets then other trends over the last two decades suggest further convergence of the two. Fully 82% of Spanish households (15 million households) now live in accommodation they own outright or are purchasing. It is clear therefore that home ownership has come to be a significant cultural phenomenon (albeit driven by processes relating to the structure of the economy and policies in the housing field) but, given that purchasing one’s home is becoming almost universal, it seems logical that, in contemporary Spanish society, attitudes to purchasing or acquiring a second home would also be viewed favourably. For example, the ownership of a primary home may be the principal asset of a household and act as surety against which other forms of purchase and consumption take place (BBVA 2005, p. 24)—including a second home. In other circumstances, the release of equity in a primary home may prompt the purchase of a second home. Indeed, some researchers (Paris 2006; López Colás et al. 2007) have suggested that second home acquisition is now an integral part of many household investment strategies. In addition, research on the motivations for acquiring a second home has shown that a key decision making factor is the intention to use a second home as a stepping-stone towards permanent retirement residence (Williams et al. 2004). In these and other ways the market for second homes may be directly affected by circumstances in the primary housing market.

Other more recent housing market processes manifest at the local scale also demonstrate the linkages between primary and second homes. “The purchase of second homes in cities reinforces the idea that leisure second home markets overlap seamlessly with housing markets” (Paris 2006, p. 9). But the growth of urban second homes relates to much more than leisure. Similar to, although somewhat later than, many other urban areas in the developed world (Vázquez 1992; White 1984), several Spanish cities have experienced gentrification processes in the recent past. A common feature in such areas is the growth of a distinctive segment of the housing market which consists of purchase primarily for investment (sometimes by institutions), acquisition for business lettings and second homes (Visser 2004; Cameron 2006). Within the last two decades a number of specific events have prompted significant property development and residential renewal in several Spanish cities, most obviously the Barcelona Olympics (Muñoz 2006), the Seville Expo and the nomination of Madrid as European City of Culture, all in 1992 (Garcia 1993). These new models of urban regeneration have been followed elsewhere, for example Bilbao (Vicario and Mártinez Monje 2003). As a consequence, a number of inner areas of such cities have experienced significant change in their socio-economic composition and their residential environments (Garcia 1993), changes which have produced identifiably gentrified neighbourhoods (Vázquez 1992). The existence of a larger number of second homes is an associated phenomenon. For example, for the Madrid metropolitan area Valenzuela (2003b) has shown how the city of Madrid itself experienced the highest rate of growth in second homes, over 50% in the 1991–2001 period. Such growth is associated with the purchase of second homes as investments (Cabré and Módenes 2004), and for use by the offspring of wealthy families and is clearly associated with the ‘gentrification’ of certain barrios (Vázquez 1992; Camarero Bullón 1997).

The number of ‘urban’ second homes has therefore grown substantially in the 1991–2001 period in several Spanish cities, for example, by 126% in Bilbao, 135% in Barcelona, 150% in Sevilla and 59% in Málaga. Not all this growth is in the central parts of these cities, of course, but an examination of the 2001 census maps (INE 2004) shows the very large numbers of centrally located second homes in these cities, for example La Vieja in Bilbao, El Raval in Barcelona, Macarena in Sevilla, Chueca in Madrid and Pizarrales in Salamanca.

Whilst the initial major process in the growth of second homes in Spain since the 1970s related to rural-urban migration and the retention of land and property in the village of origin, it is clear that, over the past four decades, a much more complex set of processes have operated. Although varied in origin and impact, it is these processes that have served to create increasingly closer relationships between first and second homes in Spain.

9 Conclusion

This examination of second homes at different scales and in different contexts demonstrates their pervasive nature and their close association with other features of the wider housing market. The weight of evidence available points clearly in the direction of a close relationship between primary and second home ownership in Spain, this being especially the case for indigenous Spaniards. Some of the theoretical reasons suggested for a dichotomy appear to be less than robust and empirical evidence at different spatial scales and relating to different features of the housing market also favours an interpretation of inter-relationship rather than independence.

Several theoretical reasons have been forwarded in favour of a contention that the market for second homes is more volatile than that for mainstream housing. One suggestion is that second homes are more likely to be sold in economically difficult periods but this argument is undermined by the fact that, in Spain as elsewhere, the distinction between first and second home has become less clear with, for example, the latter being acquired with a view to it ultimately becoming the primary home. Second, the suggestion that lenders are more concerned about defaults on second homes and are therefore less likely to lend is simply not borne out by the easy availability of credit for all purchases in the recent past. A third argument is that, in periods of depression, house price falls in mainstream housing may allow a formerly unaffordable purchase by first time buyers but this ‘cushioning’ feature is absent from the second home market. The assumption here is that second homes are a luxury and expendable commodity rather than a significant component of overall investment strategies. We have seen that, increasingly, this is not the case. Furthermore, in downturns, second homes may also actually become more affordable for some. Finally, there is the suggestion that developers are more likely to over-estimate demand for second homes, leading to the necessity of selling at reduced prices but when supply becomes limited, high prices result. The main problem with this argument is that Spanish developers do not necessarily build exclusively for either primary or secondary markets. Similar houses on the same development may be permanent dwellings or holiday homes. In this sense, therefore, the market is perceived as a unity rather than as two discrete components.

Turning to the empirical evidence, it has been demonstrated that second homes are not significantly differentiated from primary homes in terms of size of settlement or type of dwelling. Their cyclical pattern of development, in recent times at least, has not greatly differed from that of primary dwellings. The spatial distribution of second homes at the provincial scale is significantly widespread, reflecting a ‘layering’ of different processes over many years. However, it is clear that foreign ownership of both primary and secondary homes forms a major and growing segment of the Spanish housing market and the relationships between first and second home ownership within this sub-market may well be different from those in the indigenous sub-market. It is certainly evident that they need researching more closely. That is not to say that the indigenous sector of the housing market has not been subject to change, for clearly it has. But more recent processes relating to the rapid growth of owner occupation more generally, gentrification, appreciation in the value of the dwelling as an asset, and the ready availability of credit have created a ‘new’ context in the indigenous Spanish housing market which appears to be operating alongside, rather than replacing, the more traditional socio-cultural factors. But just as the ‘traditional’ processes, stemming from rural depopulation, actually created important linkages between the ‘new’ primary dwelling and the rural dwelling retained for leisure and familial reasons, so too have the more contemporaneous processes strengthened older, and created new, linkages between first and second homes. The latter are now an integral part of the overall Spanish housing market.