Airbnb, buy-to-let investment and tourism-driven displacement: A case study in Lisbon

Through detailed empirical analysis of a central area of Lisbon, this paper explores whether short-term rental platforms such as Airbnb channel investment in residential real estate and the way in which the local community is affected by the proliferation of apartments rented to visitors. Between 2015 and 2017 we conducted fine-grained fieldwork in the historical neighbourhood of Alfama and identified both the producers and socio-spatial consequences of short-term rentals. Our research did not find evidence of a sharing economy. Rather, it found a process of buy-to-let investment in which different players make profits from rents and displace residents with tourists. The paper develops two main arguments: first, we suggest that Airbnb acts as an instrument that contributes to the financialization of housing. Compared to the traditional rental market, short-term rentals offer a number of benefits that enhance market efficiency for property owners, making them increasingly attractive for both local and global investors. We found that the suppliers of short-term rentals are primarily investors who use housing as an asset to store capital. The main advantage of the short-term rental market for investors is that while they can make profits by renting properties to visitors, they can also sell them tenant-free at any moment. Second, Airbnb gives way to a hyper-flexible rental market that for tenants implies increasing insecurity and displacement concerns. We portray Airbnb as an example of buy-to-let gentrification that is experienced by residents as a process of social injustice.


Introduction
Airbnb has become the leading provider of short-term rentals (STRs), and its rapid growth is one of the most discussed topics in cities across the world. The company is certainly not the sole STR service, but it is undoubtedly the principal one. 1 As the platform is the dominant force in this market, scholars interested in exploring the proliferation of STRs have undertaken quantitative and spatial analyses aimed at mapping the supply of flats listed on Airbnb alongside measuring how much housing stock has been converted into accommodation for visitors (Adamiak, 2018;Arias-Sans and Quaglieri-Dom ınguez, 2016;Gurran and Phibbs, 2017;Sch€ afer and Braun, 2016;Wachsmuth and Weisler, 2018;Yrigoy, 2018). These studies find that STRs are mainly concentrated in city centres and tourist hotspots, and that, in such areas, a large proportion of Airbnb listings are entire apartments rented to tourists permanently rather than occasional home-sharing as Airbnb claims. In relation to this, authors have linked Airbnb with gentrification and have suggested that landlords change their apartments from long-term rentals to STRs, meaning that in the process they replace tenants with visitors (Cocola-Gant, 2016;Wachsmuth and Weisler, 2018;Yrigoy, 2018). While this conversion from residential to tourist uses is inherent to Airbnb-led gentrification, we know very little about: (a) who the Airbnb landlords are; and (b) how local communities are affected by the growth of STRs. These two issues are addressed in this paper through a case study of Lisbon. Rather than mapping the supply of STRs, between 2015 and 2017 we conducted fine-grained fieldwork in the historical neighbourhood of Alfama and identified, block by block, the producers of STRs, and explored how residents experience the proliferation of flats that are rented to visitors.
Our research did not find evidence of a sharing economy. Instead, it found an intense process of buy-to-let investment and professionalization in the provision of STRs that are driving an aggressive transformation of the area -meaning that Alfama is changing from being a working-class residential place to becoming a tacit tourist district. We portray Airbnb as an example of buy-to-let gentrification (Paccoud, 2017) in which investors purchase properties and obtain returns through the STR market, displacing residents with tourists. There are two interwoven issues here: buy-to-let investment and the socio-spatial consequences of the process. In this respect, two main arguments are developed: first, we suggest that Airbnb acts as an instrument that contributes to the financialization of housing. By the financialization of housing, we refer particularly to the housing sector's ability to absorb surplus capital (Aalbers and Christophers, 2014;DeVerteuil and Manley, 2017;Fernandez et al., 2016). In this regard, Airbnb plays an important role in a process described by others (August and Walks, 2018;Beswick et al., 2016;Fields and Uffer, 2016;Janoschka et al., 2019;Paccoud, 2017;Wijburg et al., 2018), whereby rental housing has become a new financial asset in the wake of the 2008 economic crisis. We argue that Airbnb and the consequent rise of the STR market make investment in rental housing ever more flexible and efficient. When compared to the long-term rental market, the paper highlights how digital platforms and technologies (Fields, 2019;Rogers, 2017;Shaw, 2018), STR management companies, and the inherent flexibility of the STR market offer several benefits that expand possibilities for accumulation. The main advantage of the STR market for investors is that while they can make profits by renting properties to visitors, they can also sell them tenant-free at any moment. This flexibility is key for both local and global investors who use housing as an asset to store surplus capital.
Second, we argue that Airbnb-led investment in Lisbon involves a process of buyto-let gentrification. Capital investment in a more transient, consumption-oriented built environment is displacing an existing working-class community. Buy-to-let gentrification (Paccoud, 2017) means that property owners profit by targeting more affluent tenants, which has also been called 'rental gentrification' (August and Walks, 2018;Van Criekingen, 2009). Airbnb replicates this process as tenants are replaced with visitors, fostering further housing insecurity for renters. However, we also found that homeowners who sell their flats to buy-to-let investors are an important driver of the out-migration of residents seen in the area. Additionally, the supply of apartments available in the traditional private rental market is disappearing, provoking exclusionary displacement as new residents are unable to move in. The result is a tacit change in land use from residential to tourist uses that residents experience as a process of social injustice.
The paper is organized in seven sections. We start by suggesting how Airbnb has created an additional instrument to invest in residential real estate and, in so doing, we link Airbnb with housing financialization. Next, we consider the expansion of STRs in dialogue with the literature on tourism-driven gentrification and displacement. Following these two theoretical sections, we offer a contextualization of Lisbon and show how the state plays a crucial role in promoting both tourism and investment in real estate. Then we present the case study and discuss the approach adopted in the collection of data. We selected a small sample of the Alfama neighbourhood and implemented fine-grained qualitative methods for a period of two years. The main body of the paper first examines the supply of STRs, and how the growing professionalization of Airbnb facilitates buy-to-let investment; second, it presents how long-term residents experience the process and reveals that both tenants and homeowners are displaced. In the paper's conclusion, we focus on the wider implications of Airbnb and tourism for gentrification and housing financialization; finally, we suggest ideas for policymakers and future research.

Airbnb and buy-to-let investment
Airbnb claims to be part of the sharing or collaborative economy, where digital platforms allow the exchange of underused goods and services, therefore potentially implying a more sustainable form of consumption (Dredge and Gyim othy, 2015). In this case, residents (hosts) would occasionally share their homes with visitors. However, in a study mapping the supply of Airbnb in 432 European cities, Adamiak (2018) finds that high levels of home-sharing occur only in places where tourism demand is low, while in major tourist destinations the tendency is towards increased professionalization of STR management. Studies point out that in central areas of cities such as Barcelona, Sydney, Berlin and New York, property owners increasingly rent out entire apartments permanently through STR platforms -meaning no one lives in the property -and some 'hosts' have multiple listings and portfolios that include dozens of flats (Arias-Sans and Quaglieri-Dom ınguez, 2016;Cocola-Gant, 2016;Gurran and Phibbs, 2017;Sch€ afer and Braun, 2016;Wachsmuth and Weisler, 2018). Furthermore, the professionalization of the STR sector is evident in the rise of commercial intermediaries that manage properties for third parties. For instance, the Hostmaker company (https://hostmaker.com) manages more than 550 properties in London. This suggests that, beyond some kind of sharing, in places with strong tourism demand Airbnb allows the development of new opportunities for professional actors to extract profits from real estate.
Despite the evidence of increased professionalization, we know very little about who the Airbnb landlords are. Authors have suggested that as STRs offer new potential revenue, landlords simply change their units from long-term residential leases to STRs (Cocola-Gant, 2016;Wachsmuth and Weisler, 2018;Yrigoy, 2018). However, this localized preference for short-term renters does not fully explain how flats listed on Airbnb are provided. We suggest that Airbnb opens up opportunities for buy-to-let investors to purchase properties in areas of high tourism demand and rent them out in the STR market. The argument here is that Airbnb offers a new mechanism for the financialization of housing because it provides an additional instrument for making investment in residential real estate ever more flexible and profitable. Existing studies have highlighted the growth of capital investment in private rental markets in the aftermath of the 2008 financial and economic crisis (August and Walks, 2018;Beswick et al., 2016;Fields and Uffer, 2016;Janoschka et al., 2019;Paccoud, 2017;Wijburg et al., 2018). These authors suggest that since the post-2008 capitalist restructuring, investors have implemented novel strategies that lead to the substitution of housing financialization based on the buy-to-sell model with investment in the buy-to-let market. Against this background, we argue that Airbnb, and the consequent rise of the STR market, plays an important role in this process, particularly because they have created an increasingly flexible and productive business model that accelerates the incorporation of rental housing into global circuits of capital.
Compared to the long-term rental market, Airbnb provides a set of advantages and benefits that makes the STR market increasingly attractive for both local and distant investors. We argue that three key interwoven factors are in place: platforms and digital technologies; STR property managers; and the flexibility of the STR market. First, authors such as Shaw (2018), Fields (2019) and Rogers (2017) suggest that the development of digital platforms and technologies offers several benefits that make residential real estate markets increasingly efficient, productive and available at a global scale. Following Srnicek (2017), Shaw (2018) sees Airbnb as part of platform real estate -that is, digital infrastructure connecting users in an unprecedented way. He emphasizes that there is no need to use terms like the sharing economy because digital infrastructure simply creates more efficient ways for property owners to connect with renters, thereby allowing a more productive and efficient exchange of these properties. By means of Airbnb, landlords easily reach a global market that is willing to pay high rents for short stays, making this sector increasingly profitable. Additionally, for property owners, Airbnb significantly reduces production costs. Guttentag (2015) notes that hosts have free access to services such as marketing, reservation and payment systems, which enable them to compete effortlessly with traditional accommodation enterprises. Furthermore, in the effort to cut costs, Airbnb has created unregulated marketplaces that avoid tax obligations and promote highly flexible and insecure forms of employment (Martin, 2016;Srnicek, 2017).
Second, the rise of professional operators that manage STRs on behalf of third parties additionally makes investment in this market more efficient and easier. On the one hand, STR property managers take care of the daily management of the property from marketing to cleaning and receiving guests. On the other hand, agencies that manage STRs are platforms themselves that collect daily data about the market and process it to advise property owners and investors. Furthermore, these intermediaries use owner management platforms that are able to provide distant landlords with day-to-day information about the property, allowing them to monitor their businesses in real time. As suggested by Fields (2019), the rise of digital technologies in real estate management enables a more efficient control of property portfolios, and then encourages further investment into the rental housing market. Finally, agencies provide advice regarding local regulations and legal processes and, therefore, they connect global investors with local markets.
Third, the instrument provided by the STR market significantly relies on its flexibility, which offers property owners a set of advantages compared to the long-term rental sector. A first point usually suggested by authors is that landlords make more profits in the STR market (Wachsmuth and Weisler, 2018;Yrigoy, 2018). Additionally, we suggest that Airbnb gives way to the rise of a hyper-flexible rental market that avoids the restrictions that longterm rental markets have. What is particularly important is that Airbnb allows landlords to repossess properties at any moment, which does not happen easily in regulated long-term rental markets (Paccoud, 2017). Airbnb allows investors to make profits from rents, but without having permanent tenants using the property, meaning they can sell the property or use it as they will. For investors seeking to store their capital in residential real estate, Airbnb allows them to sell the property free of tenants at any time, meanwhile making profits from rental housing. Additionally, there is evidence to suggest that second-home buyers are key buy-to-let Airbnb investors (Montezuma and McGarrigle, 2019). The flexibility of Airbnb gives these individual investors the ability to earn returns on their properties, renting them out on Airbnb while they are not in use. The rise of professional operators that manage Airbnb properties for distant landlords further facilitates this process. Finally, high turnover rates allow landlords to increase rental prices (Paccoud, 2017;Van Criekingen, 2009); however, in Airbnb, this rent price speculation can occur on a weekly basis. This increases the expectations of obtaining more revenue than longterm rentals.
In conclusion, we suggest that Airbnb has created a new opportunity to further advance financial investment into the rental housing market, enabling the arrival of buy-to-let capital into certain tourist destinations. The platform reinforces the neoliberal paradigm (Martin, 2016) because it enhances market efficiency for owners of real estate capital and gives way to an increasingly flexible rental market that, for tenants, implies further insecurity and displacement concerns.

Airbnb and tourism-driven displacement
Several authors have noted that the most significant consequences of the post-2008 financialization of rental housing are: (a) a progressive substitution of local property owners by transnational corporate investors and globalized landlords; and (b) a process of rental gentrification whereby these corporate investors target higher-income renters, who can guarantee higher returns (August and Walks, 2018;Beswick et al., 2016;Fields and Uffer, 2016;Janoschka et al., 2019;Paccoud, 2017;Wijburg et al., 2018). We argue that Airbnb-led investment contributes to this process: investors replace local property owners and pursue higher returns by replacing long-term tenants with tourists. The replacement of tenants with short-term visitors is a form of 'direct displacement' (Marcuse, 1985), that is, residents move involuntarily from a residence. Different regulatory frameworks may influence this process because landlords can easily terminate contracts and rent their flats on platforms such as Airbnb if tenants have insecure tenancy agreements. The state thus plays a key role in advancing or limiting the process.
However, Airbnb's social impacts go beyond the direct displacement of tenants. It has been noted that homeowners living in tourist areas in cities such as Berlin, Amsterdam and Barcelona are 'voluntarily' selling their flats (Cocola-Gant, 2016;Fu¨ller and Michel, 2014;Pinkster and Boterman, 2017). Buy-to-let investors searching for properties to rent out on Airbnb target homeowners living in tourist areas as well as local landlords. The pressure of buy-to-let investors searching for properties, together with the everyday disruptions caused by visitors in such areas, leads to a situation in which selling the flat and moving out is an appealing option for residents. While the decision to sell the family flat may be voluntary, it reinforces a process of social change in which permanent residents are replaced by transient users. Nonetheless, there is evidence to suggest that homeowners selling their flats in tourist areas do so less to take advantage of better opportunities, and more so with resignation, because tourism disruptions affect their health and daily well-being (Cocola-Gant, 2016;Pinkster and Boterman, 2017). As these authors suggest, such relocations constitute a loss of place, in which the daily pressure of tourism leads people to move out.
Finally, the touristic use of housing stock severely reduces the supply of apartments available for long-term occupation to such an extent that people are unable to find units offered in the private rental market. For instance, Sch€ afer and Braun (2016) found that in some streets of central Berlin all of the available flats are let out to tourists. In a more striking case, Ibiza faces a situation in which workers such as teachers and doctors refuse to work on the island because they cannot find a house to live in (Colmenero, 2016). The result is that as people move out, with new residents unable to move in, the spread of STRs may lead to the conversion of some residential areas into tourist districts. Therefore, the proliferation of STRs can foster tourism gentrification as a process of land use change from residential to commercial uses in which transient visitors replace long-term residents (Cocola-Gant, 2018;Gotham, 2005).

Lisbon: the role of the state in promoting real estate investment and tourism
Lisbon is the capital of Portugal and has a population of 547,733 inhabitants. As in other Southern European cities, Lisbon's deindustrialization and transition towards a post-industrial economy, which started in the 1970s, focused significantly on leisure. Since the 1990s, local urban revitalization strategies aimed to place Lisbon on the map of international tourism and to attract global real estate investment (Barata-Salgueiro et al., 2017;Malet-Calvo, 2018). The 2008 financial crisis further accelerated this strategy as the attraction of both tourism and foreign investment came to be seen as a solution to the recession (Mendes, 2018). Lisbon has been positioned as an ideal destination for visitors and international students, digital nomads and other transnational mobile populations in search of an exciting and cheap place to live in Southern Europe. The government has also targeted second-home buyers to attract foreign investment (Montezuma and McGarrigle, 2019). In 2009, the Non-habitual Residents programme was introduced which offers low personal taxation to EU citizens. This programme stimulated the arrival of Europeans, particularly retirees. Similarly, in 2012, an immigrant investor initiative known as the Golden Visa programme was approved. This offers freedom of movement within the Schengen area and future Portuguese citizenship to non-EU applicants who buy properties for at least e500,000 or invest e350,000 in property for rehabilitation.
The state implemented further policies to make the local real estate market more attractive to global capital, such as offering fiscal incentives to real estate investment funds and speeding up licensing proceedings for them (Barata-Salgueiro et al., 2017;Mendes, 2018). In addition, Portuguese banks were rescued by the European Union in 2011, which involved not only austerity measures but increased flexibility in the real estate sector. Particularly important was the New Urban Lease Law passed in 2012 that deregulated the private rental market. The new law made it easier for landlords to end tenancy agreements. Furthermore, there is no longer a minimum length for new rental contracts; previously the requirement was for at least five years. The flexibilization of tenancy agreements made the rental market particularly attractive to private capital.
As a result, Lisbon is experiencing a significant growth of tourism and real estate investment, the latter focusing heavily on housing rehabilitation in run-down central areas (Barata-Salgueiro et al., 2017;Lestega´s, 2019). Rehabilitation is notably marked by the opening of commercial services for visitors, from hotels and STRs to restaurants and nightlife activities (Nofre, 2013). For instance, the number of hotels doubled between 2008 and 2018, from 105 to 218. Additionally, rental prices grew by 110% over the period 2013(Relvas, 2018. Given low average salaries -the minimum monthly wage is e580 -the state-led strategy to attract tourism and foreign investment has contributed to a dramatic housing crisis for the local population. The state's measures have worked in tandem with the expansion of digital platforms such as Airbnb, compounding its impact on local residents. In what follows, we explore both the production and social consequences of the expansion of STRs.

Case study and methodology
The number of Airbnb listings in Lisbon dramatically increased after 2014. According to the website Airdna.co, in 2018 there were 15,700 active listings, of which 76% were entire homes. In 2014 there were just 2000 of these listings. Mapping the supply of STRs is useful to provide a broad picture of the phenomenon, but it tells us very little about investment in the housing market, who the Airbnb landlords are and how STRs transform urban communities. To learn more about this, we designed a fine-grained qualitative case study of part of the Alfama neighbourhood in Lisbon (Figure 1).
Alfama is the oldest part of the city. In the Middle Ages, it was built on the hill that extends from the castle to the Tagus river. Alfama has been a traditional tourist attraction in Lisbon mainly due to its Moorish-rooted architecture, alleyways, central location, proximity to the castle and its links to Fado -a famous type of Portuguese folk music (Firmino da Costa, 2008). However, prior to the 2008 financial crisis, tourism was still an emerging industry and there were no hotels or holiday rentals in the neighbourhood. Alfama was a working-class residential area. According to the 2011 census, the population was 3042, with a high rate of lower-educated, 2 older residents: the population aged 65 or above represented 28% of residents, and was the main age group in the area; only 18% of the adult population had university degrees (in Lisbon this rate was 32%), while 35% was educated to primary school level. Furthermore, the 2011 census indicates that more than half of residential buildings were in a state of disrepair and 10% were severely degraded. Slightly more than three-quarters (77%) of occupied dwellings were in the private rental market. In sum, Alfama was mainly inhabited by working-class, elderly tenants, living in a decaying built environment. Prior to the New Urban Lease Law of 2012, tenancy agreements could last for a lifetime, meaning that elderly residents who were born in the neighbourhood are still present in the area. As Firmino da Costa (2008) illustrates, it is an area with a strong sense of community where 'everybody knows each other' due to long-term tenancy and social bonds created over time.
The case study area selected for this research comprises the main street of Alfama (Rua dos Remedios) and its surroundings, an area of 3.6 hectares (Figure 1). There are 251 apartment blocks in the area and the housing stock comprises 945 units. We studied the area for two years, from 2015 to 2017. In 2015 the neighbourhood was undergoing a rapid rehabilitation process and there was a noticeable growth of services catering to visitors, such as restaurants, pubs with staged Fado music and especially holiday rentals. We were interested in identifying the extent to which the area's housing stock was used as STRs; who the landlords of the STRs were; and whether there were links between STRs and rehabilitation of housing stock, home sales and displacement. We divided the research design into two phases. First, we surveyed the area block by block in December 2015 and again a year later. We visited every apartment block and collected information through informal interviews with residents living in the blocks, shop keepers, real estate agents, landlords and STR property managers. We also lived in the area and this was crucial to gaining access to the community. For instance, residents and shop keepers provided information about tenants that were displaced in order to open STRs, rehabilitation works, landlords who sold units to investors and whether flats were rented to visitors permanently or temporarily. Airbnb property managers working in the area gave us details about owners of STRs. In all cases, the information gathered was provided by at least two informants. This in-depth field exploration was possible partly due to the size of the case study area. Furthermore, the emergence of tourism and Airbnb is a daily topic of conversation in the neighborhood and people were very open and willing to talking to us. In sum, the focus on a small area and being there on a daily basis allowed us to monitor social and spatial transformations in a very detailed way. In particular, by surveying the 251 blocks, we were able to identify: the supply of STRs; the type and availability of STRs (entire flats or single rooms and whether the apartment was available on Airbnb permanently or temporarily); the Airbnb landlords (whether they were residents sharing their homes, landlords that stopped renting to tenants or investors); flats that have been rehabilitated; units that have been purchased by investors, and type of investor (individual or corporate); the previous use of the flat (whether it was vacant or used by residents); and, finally, we were able to map the displacement of tenants. Additionally, we verified the results of the survey by analysing the Short-Term Rental National Register (Registo Nacional de Alojamento Local), which provides information about the addresses and owners of STRs. 3 Second, we conducted 25 in-depth interviews, of which 15 were with residents and 10 with real estate market key informants. We asked residents about how STRs were transforming the area and how they experienced this transformation. As other authors suggest (DeVerteuil, 2011;Newman and Wyly, 2006;Slater, 2006), qualitative research is useful to help better understand the processes of displacement as they relate to gentrification because a simple measurement of displacement leaves important aspects of urban change unexplored. Key informants were real estate agents, owners of STRs and agencies that manage STRs. These interviews focused on Airbnb and investment in the housing market. STR property managers had portfolios between 70 and 200 flats across Lisbon and their views were central to understanding the strategies of Airbnb investors.

Airbnb in Alfama: professionalization and buy-to-let investment
According to our research participants, the first apartment rented to visitors in the case study area appeared in 2011. In December 2016, we found that 25% of the entire housing stock consisted of STRs, accounting for 235 apartments out of 945. A year earlier, STRs formed 16% of the housing stock. In addition to this rapid growth, we want to highlight four key interwoven findings. First, among the supply of STRs only two apartments were homes shared by residents. In other words, 99% of the supply of STRs consists of entire apartments that are marketed all year round and belong to people who do not reside in the neighbourhood. In fact, in this small area, 14 apartment blocks are used entirely as STRs, which makes them de facto hotels. This suggests that there is no evidence of a sharing economy. Rather, we found a rapid change in the use of the built environment in which housing is now used as tourist accommodation.
Second, this tacit shift in land use has come on a wave of real estate investment and housing rehabilitation (Figure 2). We found that the main investment strategy is buy-to-let in the STR market. A second investment strategy is 'buy-to-leave' (Fernandez et al., 2016). During the period of two years, we found that 150 apartments were purchased in the study area; only one was purchased by a new resident. Most of the others were used as STRs and a few apartments remained vacant. Interestingly, we did not identify any instance of investment in rental housing focusing on the long-term residential market. The incorporation of capital into rental housing is entirely focused on the STR market in this case. The area is also experiencing an intense rehabilitation process that is following a similar pattern in which rehabilitated buildings are either turned into STRs or apartments remain vacant. In the study area, 21 apartment blocks with 130 flats were rehabilitated during the 2015-2017 period. Among the blocks, 12 were converted entirely to STRs and are therefore de facto hotels.
Third, the fine-grained fieldwork allowed us to identify the Airbnb landlords in 2015 and 2016 (Figure 3). We found four types of Airbnb suppliers. The first are residents sharing their homes. As stated, in 2016 only two apartments correspond to this category. The second is homeowners that moved to different areas of Lisbon and rented their flat in Alfama to visitors. In 2016 we found nine of these cases. The third category consists of local landlords who ceased renting to tenants and listed their flats on Airbnb. In 2016, 40 apartments representing 17% of the supply of STRs fit in this category. However, in 2015 local landlords supplied 32% of STRs. The rapid rise in the fourth category has eclipsed all others: buy-to-let investors. In 2016 this category offered 184 apartments, representing 78% of the supply of STRs. The tendency is that buy-to-let investors gradually replace both local landlords and homeowners. In relation to this, in Figure 3 we refer to individual investors as people who purchased one or two flats and corporate investors as companies that bought entire apartment buildings and converted them into accommodation for visitors.
Individual investors include national and foreign investors; however, the latter are replacing the former. Our findings confirm Montezuma and McGarrigle's (2019) results that immigrant investors and owners of second homes play an important role in buying properties that are let out via Airbnb. According to Airbnb property managers, foreign investors are not very wealthy elite, but middle-class individuals who have far fewer 'disposable assets' than the super-rich (Rogers and Koh, 2017). In particular, these investors are from France, the UK, Brazil and China, and are attracted by the fiscal benefits given to them by the government. The STR market allows second-home buyers both to use the house when they are in Lisbon and to obtain revenue when they are not. Therefore, Airbnb is creating a new mechanism that further emphasizes the links between the mobility of the transnational middle-classes and their investment in overseas homes (Hayes, 2018;Montezuma and McGarrigle, 2019). Corporate investors include national and transnational players that buy entire buildings to refurbish and rent the units on Airbnb. We identified national firms that have a long presence in the local real estate market, and that have expanded their business strategy in recent years by investing in this new sector. The arrival of global capital is also evident in the area. Some buildings have been refurbished by French developers who display their logos on the fac¸ade. However, we were not able to discover the national origin of capital in many cases. Informants such as STR property managers mentioned that the business was led by corporate foreign investors, but they either did not have more details or did not want to disclose them, therefore highlighting the challenges that researchers face in terms of gaining access to the practices of global investors and the super-rich (Rogers and Koh, 2017). In any case, Figure 3 shows that corporate buy-to-let investors are the main group of Airbnb landlords, which is also explained because they acquire entire buildings rather than single flats. The tendency is a gradual change from being a business supplied by local landlords and individual investors to a commercial sector dominated by sizeable developers.
Finally, a fourth important point is the professionalization of the STR sector and how Airbnb provides a flexible instrument that brings more benefits to investment in residential real estate. Services such as cleaning and receiving guests were first done by landlords, but it has rapidly become a more organized activity provided by agencies. Interviews with these property intermediaries reveal that they manage holiday rentals for both local landlords and foreign investors, but as one property manager suggests, 'the majority of our clients are foreigners who do not live in Lisbon. This is why they need a local agent.' In this regard, the professionalization of the sector is further facilitating buy-to-let investment. Such companies take care of the entire commercialization process, from marketing and reservation to receiving guests, and they advise foreign investors on how to buy properties in the area, list the flats on Airbnb, and manage the properties on their behalf. We found that STR property managers have partnerships with law firms and local real estate agents to assist foreign investors throughout the entire process, from legal advice to finding properties in convenient locations, refurbishing them and, finally, commercializing the apartments on Airbnb. While local investors are more likely to use their networks in Lisbon to buy run-down apartments owned by non-professional landlords, foreign investors rely on intermediaries to access the local market, meaning that agents are central to allowing the flow of international capital into the STR market.
Property managers explained that the advantage of the STR market is not only that investors obtain more revenue than with long-term rentals, but that the platform adds a new incentive for those who use the housing market as an asset in which surplus capital can be stored (Aalbers and Christophers, 2014;DeVerteuil and Manley, 2017;Fernandez et al., 2016). The benefits of Airbnb rely on the freedom that it provides to owners of real estate capital, especially because they can make profits with no tenants in the property. One property manager explained: The investment is to buy a property, leave the money there, and sell it eventually. Investors normally leave properties vacant; they do not want tenants because you sell better if the property is vacant. But Airbnb adds something new to this kind of investment. They make profits from rents but free of tenants. They can also use the flat if they are in Lisbon. We keep the property really clean and nice for the tourist and this is another benefit for the investor: they will sell it better if the flat is elegant and well maintained. This flexibility of the STR market is, according to a property manager, 'the perfect combination'. Owners of second homes obtain returns by renting the properties out on Airbnb, yet the nature of the STR market allows them to use the property when they visit Lisbon (Montezuma and McGarrigle, 2019). Investors using housing as an asset in which money can be stored can additionally make profits from rents. For tenants, however, the flexibility of Airbnb means further insecurity. We focus on the experiences of residents below.

Experiencing neighbourhood change: tourism, Airbnb and displacement
We are invaded! Invaded! And property owners just want money. They do not care about human rights. Someday we will walk around here and say: 'Look! This sort of hotel used to be my home!' This is going to be our history. (Female resident) During the interviews with residents, a recurrent topic was direct displacement and a consequent substitution of tenants with visitors. Authors have stressed the methodological struggles of quantifying direct displacement since displaced residents typically disappear from 'the very places where researchers and census-takers go to look for them' (Newman and Wyly, 2006: 27). However, the intensive field research that we implemented allowed us to identify and map specific cases of direct displacement (Figure 2). Over a period of two years, we identified 27 apartments of which 36 tenants had been forced to move out. Among these cases, 18 apartments gave way to STRs and 9 remained empty. Moreover, stories about residents at imminent risk of displacement are commonplace, making it likely the number will continue to rise. Landlord strategies to displace tenants mirror other examples of gentrification in which they pressure existing residents to force them to move (Marcuse, 1985;Slater, 2017). In the wave of rehabilitation that the area experiences, a common strategy used by property owners is to renovate the building to open STRs, yet the flats occupied by tenants remain untouched. Despite the physical improvement of the housing stock in the area, this practice of selective rehabilitation means that there were no benefits for tenants even though they had to suffer from the noise and disruption of the work taking place in the building for several months. Furthermore, residents experience landlords' practices as a form of harassment, which leads to a sense of helplessness and frustration. As noted in other cases (DeVerteuil, 2011;Marcuse, 1985;Newman and Wyly, 2006;Slater, 2009), residents who remain often do so because of adaptive survival strategies amidst a general degradation of their quality of life. A low-income elderly tenant with no other options explains that: They rehabilitated the other flats, but they flooded ours. They said they would fix it. They removed the beams and did not replace them. And now they sold the building and left us like this. They never came to fix it! The strategies used to harass renters in which apartment blocks are renovated while ignoring repairs in units occupied by long-standing tenants have also been documented in multifamily apartments undergoing financialization (August and Walks, 2018). Other authors further show that corporate landlords harass tenants on a regular basis in order to promote direct displacement so as to profit from tenant turnover with higher rents (Fields and Uffer, 2016;Janoschka et al., 2019). As stated, the spread of Airbnb is replicating the strategies seen in the financialization of rental housing. Whether the refurbished units are for higherincome renters or Airbnb visitors, buy-to-let gentrification implies the direct displacement of low-income residents as they are seen as a barrier to capital accumulation.
So far, we have referred to cases of direct displacement of tenants triggered by the pressure applied by property owners to force them to move. It is also important to note that the area is experiencing a process of out-migration and social change inextricably linked to tourism and Airbnb buy-to-let investment. As noted in other tourism destinations (Cocola-Gant, 2016;Colomb and Novy, 2016;Pinkster and Boterman, 2017), residents move out because the tourism-led transformation of the neighbourhood makes everyday life increasingly difficult. This process has been highlighted by several of our participants: I am from Alfama but living here has become hellish and this leads to the depopulation of the area. Everyday life has become impossible here. Commercial change; overcrowding, we do not have physical space to walk; traffic congestion; noise; dust from the rehabilitation works.
We identified cases of tenants that decided to move due to this tourism-led transformation. Other tenants left the neighbourhood for personal reasons. In either case, landlords took the opportunity to list the apartments on Airbnb rather than renting them to other long-term tenants. Most commonly, homeowners sell their apartments to new investors, who then rent them out in the STR market. In a context of excessive tourism growth and annoyances caused by rehabilitation in apartment blocks, property agents asking residents to sell their flats is a common practice in the area, and many people saw this as an appealing option. The truth is that among the 235 STRs, 74 were vacant (degraded apartments with no residents), and 68 were occupied by tenants and 93 by homeowners. In sum, tourism-driven displacement affects more than just poor tenants. It undermines the role of neighbourhoods as a space of social reproduction and, in such contexts, residents 'voluntarily' leave.
Therefore, in understanding how displacement takes place, we need to explore the daily experiences of people living in gentrifying areas, rather than simply measuring how many residents have been spatially displaced (Davidson, 2009;DeVerteuil, 2011;Newman and Wyly, 2006;Slater, 2006). In this case, we found that residents who remain experience both the pressure from landlords and neighbourhood change as a process of social injustice. The neighbourhood is becoming a dual space in which renovated flats are used by visitors and degraded ones are occupied by local residents. An important question that emphasizes the frustration of residents is the disintegration of community relations due to the outmigration of residents and their substitution by transient users. As residents are moving out, apartment buildings are being used by visitors who do not speak the local language, do not mix with local residents, and whose routines are marked by a touristic desire to have fun detached from any obligation to others in the local community.
Finally, it is worth noting how STRs cause exclusionary displacement (Marcuse, 1985;Slater, 2009). In 2017, over a period of 6 months, we monitored the Idealista website weekly -the main website for property rentals in Portugal -and found that in general there are no apartments to rent for long-term occupation. On occasion, there were one or two flats available, and during one investigation we found three, meaning that the rental market is almost fully oriented towards visitors. Additionally, as the local population cannot afford the few flats that are available for long-term occupation, people who occupy such apartments are usually privileged transnational migrants and international students. The lack of accommodation for long-term occupation is a central concern for the community. People who are at risk of direct displacement are forced to accept poor living conditions because they do not have other options available to them in the area. Of the 36 people who suffered direct displacement, none of them remained in the case study area and only two were able to stay near Alfama. The impossibility of finding affordable accommodation also triggers feelings of frustration: I had to leave the neighbourhood because I didn't have any other option. Otherwise I wouldn't leave! I was born here, raised here . . . I wanted to die here. But . . . [voice trembles, eyes get watery] everything is for tourism now . . . And they [the landlords] have all the power . . . If I had found another apartment here, I would have stayed.

Discussion and concluding remarks
By exploring the suppliers and socio-spatial consequences of Airbnb in Alfama, this case study reveals several points about this emerging topic, opens new questions for further research, and provides evidence to inform policymaking. Contrary to the claims of Airbnb that hosts are 'ordinary people', and that guests use such apartments to 'live like a local' (Dredge and Gyim othy, 2015;Guttentag, 2015), our findings show that hosts are investors, professional developers and landlords, while the locals have been displaced and experience the proliferation of STRs as a process of social injustice. In relation to this, an important issue is that STRs in Alfama are driving the gentrification of the neighbourhood. In fact, the sociospatial impact of Airbnb in this case replicates dynamics seen in cases of classical gentrification -that is, housing rehabilitation in a run-down central urban area and the displacement of lowincome and elderly residents (Lees et al., 2008). In a very short period, Alfama has moved from a phase of abandonment and decline to a new phase of physical upgrading, escalating property prices and social change. Gentrification in this case is not fuelled by local labour restructuring and an expanding urban middle-class as it might in the classical initial steps of gentrification (Lees et al., 2008). Rather, neighbourhood change in Alfama is driven by the extra-local demand that tourism offers and the role of Airbnb in creating new real estate investment opportunities. We showed that 78% of Airbnb landlords are individual and corporate investors and, in this regard, we agree with Aalbers (2019) that Airbnb is an integral part of the current fifth-wave gentrification in which the arrival of global real estate capital into certain neighbourhoods increasingly drives the process.
Airbnb channels the absorption of capital into the housing market and, as such, it has to be considered as an important element in the financialization of rental housing. In this case, the state plays a crucial role in advancing the process via the opening of the local market to global capital. However, Airbnb, as well as professional property managers, add a new incentive as investors are benefiting from the flexibility of the STR market to further use housing markets as an asset to store surplus capital, and in the process make profits from rents without the 'barrier' of having a tenant in the property. Furthermore, the platform makes local residential markets available at a global scale, meaning that the internet and digital platforms are technologies that speed up the globalization of real estate markets (Rogers, 2017). It is worth noting that financial institutions and the super-rich are not the only players in the process. The rise of STR property managers facilitates second-home owners' and mobile middle-classes' investment in Airbnb (Montezuma and McGarrigle, 2019). This paper is the first attempt to link Airbnb with housing financialization and therefore we need additional comparative studies. An important point should be to explore how Airbnb-led investment is connected to other investment strategies taking place in major tourism destinations. In Southern European cities experiencing tourism growth, authors have pointed out that financial institutions now see hotel chains as another market segment (Janoschka et al., 2019;Yrigoy, 2016). In addition, we showed that together with buy-to-let investment, several players undertake buy-to-leave strategies, in which apartments remain vacant. It seems that investors seeking opportunities in residential real estate do so in emerging tourist destinations as they may be encouraged by the intensification of land use in such places and consequent rises in property values (Gotham, 2005). This sort of exploration is important because residents in tourist destinations usually blame the tourist as a competitor for housing (Colomb and Novy, 2016;Fu¨ller and Michel, 2014), while it may actually be the case that local communities compete against mobile middle-classes, the super-rich and financial institutions that are helped by digital platforms to pursue new wealth-building strategies.
Airbnb buy-to-let investment is driving the displacement of an urban community, both tenants and homeowners. On the one hand, as Martin (2016) suggests, Airbnb is reinforcing the neoliberal paradigm that in this case involves further insecurity of tenancy. On the other hand, the excessive growth of tourism and investor demand motivates residents to sell their flats and move out. Additionally, the rental market caters to visitors rather than to long-term occupants, reducing housing alternatives for residents. This process reinforces the housing crisis that Lisbon faces today and, in Alfama, it results in a rapid change in land use that is increasingly making the area a tourist district, involving the substitution of a local community by transient users. In other words, the socio-spatial consequences of the proliferation of STRs pose important questions about tourism and the right to the city. In relation to this, we showed that people who managed to remain experience a sense of loss and frustration as well as daily disruption that undermines their quality of life. As DeVerteuil (2011) observes, we should consider the disadvantages of passively 'staying put', especially because the displacement literature treats immobility as inherently positive and unproblematic. We must further explore how the transient nature of visitors and their differentiated patterns of consumption accentuate the feeling that places have been appropriated by 'others'. The residents' perception of how Airbnb transforms their communities requires further detailed exploration.
Finally, our findings tell us important things about feasible regulation of the excessive growth of STRs. A central question is to note the role of the state and political decisions in advancing Airbnb and tourism-driven gentrification. The proliferation of Airbnb and tourism growth has been facilitated by the deregulation of the real estate sector; the opening of the local market to global capital; and state-led tourism promotion as a solution to deindustrialization and the 2008 financial crisis. We have, nevertheless, provided evidence that in our case study in Alfama only 1% of the supply of Airbnb properties is offered by residents sharing their homes. This is a key point for public policy. In different cities around the world, current attempts to regulate STRs assume that Airbnb listings are supplied by households taking in extra income, and this chiefly motivated policy responses that limit the number of days that residents can share their homes with visitors (Aguilera et al., 2019;Ferreri and Sanyal, 2018;Holman et al., 2018;Nieuwland and van Melik, 2018). Our findings reveal that such policies will not work to control the excessive proliferation of STRs simply because they rely on assumptions about the sharing economy that do not fit the evidence of how the STR market is formed, and who its main players are. It seems that authorities have accepted the rhetoric of Airbnb, probably influenced by the efforts of the company in lobbying governments (Aguilera et al., 2019;Guttentag, 2015). Alternative policy solutions are needed if we are serious about mitigating the socio-spatial effects of the expansion of STRs. It seems to us that they should focus on: (a) reversing the current insecurity of tenancy; (b) a taxation framework that levies short-term occupation and incentivizes long-term rentals; (c) a reform of fiscal benefits given to foreign investors and financial institutions; (d) a reconsideration of the role of tourism as a sustainable form of economic growth; and (e) an overall radical re-composition of the role of housing as shelter rather than an investment vehicle.