Abstract
Social housing providers in many advanced economies have been directed towards market-orientated and commercial business models, whereas the scope of the private rental sector has been increasingly expanded to the provision of housing for low-income households. Where these developments coincided, the demarcation between the activities of social and private landlords has blurred, with the result of increasing competitive pressure on both groups. This paper sheds light on the behavioural aspects of competition by introducing the concept of inter-landlord rivalry in local rental housing markets. Drawing on data from 36 in-depth interviews with social and private landlords in Coventry/England and Breda/the Netherlands, this study shows that existing perceptions of rivalry among most landlords are exceptional in low-income renting but pronounced in more expensive, commercial rental segments. The paper will demonstrate that these perceptions are highly subjective and non-reciprocal. Moreover, competitive perceptions and interactions are strongly affected by political and market structural settings, and appear to be dependent on the trade-off between the social mission and commercial goals of individual housing associations in the social housing sector.
Introduction
In the last two decades, there has been a surge of public housing stock transfers to private non-profit housing associations (HAs), tenant cooperatives and profit-oriented landlords in many European countries (Scanlon and Whitehead, 2007; Stephens et al., 2008). This has been accompanied by a shift in social housing finance from object subsidies to means-tested subject subsidies (Kemp, 2007), and a political imperative for hybridity in the social housing sector (Mullins et al., 2012; Rhodes and Mullins, 2009) through, for instance, the provision of commercial rental units or the development of owner-occupied housing and non-residential real estate (Haffner et al., 2009; Lennartz, 2011). Simultaneously, some governments have tried to strengthen the role of private for-profit landlords in the provision of affordable rental housing for low-income households and homeless people in various countries. Popular policy measures in that regard were the expansion of means-tested subsidies for private renters, tax breaks for the development of affordable housing units, and/or the opening up subsidy schemes beyond traditional social landlords (Kemp, 2011; Scanlon and Kochan, 2011).
In trying to make sense of the waning boundaries between the activities of social and private landlords, housing researchers have proposed the concept of competition as a meaningful approach to analyse their relation in contemporary housing markets (e.g. Atterhög and Lind, 2004; Hulse et al., 2010). What connects all of the existing approaches is that competition has been predominantly understood in terms of its structural and institutional settings, focusing on, inter alia, landlords’ respective tenure shares of social and private renting, the ease of access to the social housing industry and object subsidies, the market domain overlap of the two landlord groups, or the level of product similarity with regard to rent levels and tenants’ property rights. Good examples here are the concept of a ‘substitutability gap between social and market renting’ by Haffner et al. (2009), a model of ‘mixed rental markets’ by Lennartz (2014), and the theory of ‘integrated rental markets’ as proposed by Kemeny et al. (2005).
The key argument of this paper is that the behavioural aspects of competition between social and private landlords have, however, been left out of the picture and were just assumed to follow from structural and institutional conditions in housing markets. Yet, such an approach does not do justice to the different manifestations of competitive behaviour, thus neglecting the cognitive and relational elements of inter-landlord relations. This paper sets out to fill this gap in the housing literature by exploring the reciprocal perceptions of social and private landlords, their competitive interactions in terms of strategic decision making, and finally, the circumstances under which competitive behaviour is most likely to develop. For this, we lean on the concept of inter-firm rivalry and apply it to a qualitative study of landlord behaviour in two local rental markets in two countries, Breda in the Netherlands and Coventry in England.
Theoretical context
Competitive markets and inter-firm rivalry
A brief and arguably simplified version of the concept of behavioural competition in the mainstream economics literature is that it should first and foremost be understood as a property of the market structure in which providers operate. Behaviour here is determined by the similarity of the products that suppliers sell, the number and market shares of individual suppliers, and how easy it is to enter and exit that market. Under the assumption that all providers share the primacy of profit maximisation in the long run, each structural setup, be it perfect competition, a monopolistic structure, monopolistic competition or an oligopolistic structure implicates specific firm behaviour with regard to optimal pricing, quality and quantity outputs. The main difference between perfectly and imperfectly competitive markets is that firms in the latter may adapt their business strategies to increase their profits, while firms in a perfectly competitive market are price-takers and have no influence on optimal outputs (Martin, 2010).
A different approach to understanding competitive behaviour is the notion of inter-firm rivalry (e.g. Baum and Korn, 1996). It departs from the mainstream approach by emphasising the role and everyday decisions of firm managers and owners, focusing on individual and collective perceptions of competitors and market environments and how these develop into actual interactions in terms of pricing, quality setting and advertising. The essence of rivalry is that at least two firms are actively striving for incompatible positions in the market. Four key aspects have been noted in the literature: First, inter-firm rivalry is relational as it is based on the actions and reactions of competing firms. Competitive behaviour is thus a dynamic process, which is shaped by conscious strategic choices of managers rather than being a given exogenous fact (Kilduff et al., 2010). Second, it takes place between firms directly, which means that a firm’s competitive actions are consciously targeted at their rivals in order to gain a competitive advantage. To quote Porter (1980), ‘firms feel the effects of each other’s moves and are prone to respond to them’. Third, managers have incomplete and biased knowledge and information about markets and competitors, and they have differing cognitive capacities to understand these. As a result, there might be a cleavage between objective (as measured by structural criteria such as supply concentration) and perceived levels of competition, meaning that subjective psychological stakes can trigger individual managers to depart from rational economic behaviour (Nair and Selover, 2012). Fourth, the subjective cognitive nature of rivalry means that it is not necessarily a reciprocal perception. In reality, one firm might see another firm in the market as a close competitor, while the other might not perceive a competitive relationship at all. This can lead to rivalrous behaviour of one firm, which does however not provoke a rivalrous reaction of its alleged competitor (Chen et al., 2007). In the same vein, competitive behaviour could be targeted at one competitor in a given market, but not necessarily at another one (Chen, 1996).
Notwithstanding their conceptual differences, various studies have sought to analyse how structural conditions may shape inter-firm rivalry (for an overview see Baum and Korn, 1996; Kilduff et al., 2010). Nair and Selover (2012) find that perceptions of rivalry and rivalrous strategic behaviour are strongly associated with high levels of product similarity as well as similarly large firm size. Furthermore, they show that rivalry appears to be positively influenced by the market domain overlap of competing firms. Where firms operate in various identical market segments rather than only in a single market, the chance that they identify and direct their competitive behaviour at each other significantly increases. Paton and Wilson (2001) add to this that that rivalrous interactions intensify when the same firms have operated in the same market for longer periods, suggesting a time component of inter-firm rivalry. On the other hand, there is ample evidence that rivalry is negatively associated with a large and increasing number of small suppliers in the market. An explanation here might be that larger firms have better financial and cognitive resources to engage in such behaviour; i.e. they have sizeable marketing budgets, they have business strategy departments, or they can hire external competitive intelligence experts (Chen et al., 2007). To sum up, the concept of inter-firm rivalry opposes the mainstream competitive market theorem in that it does not increase linearly from monopolistic markets (where there is no competition and rivalry at all) to perfectly competitive markets, but it seems to be maximised in an oligopolistic structure where only a few larger providers have a large and equal amount of market power.
Inter-landlord rivalry in mixed rental markets 1
Undoubtedly, the mainstream notion of competitive behaviour as a structural contingency is a meaningful and straightforward approach to understanding conditions and outcomes of welfare-optimal and profit-maximising firm behaviour. Yet, the contention here is the concept of inter-firm rivalry is arguably a more fitting approach if one wants engage in an initial exploration of competitive interactions between the social and private landlords in mixed rental markets. After all, the literature does not provide any evidence on whether the two landlord groups actually perceive each other to be competitors and what their strategic options are when engaging in competitive behaviour.
As a result, this study applies a concept of inter-landlord rivalry, in which competitive behaviour is understood as a cognitive process where social landlords feel pressures from private landlords – and vice versa – and are prone to respond to these in their strategic decision making. Accordingly, it will be investigated whether there is a (i) relational, (ii) direct, (iii) subjective and (iv) possibly non-reciprocal form of competitive behaviour between the two landlord groups. In a subsequent step the paper uses a comparative approach to discuss the drivers of (and barriers to) rivalrous behaviour in mixed rental markets. This touches on the question of the structural circumstances under which inter-landlord rivalry is most likely to manifest itself.
A limitation in the application of inter-firm rivalry concept is that it relies on the notion of profit-oriented business models, meaning that all firms try to outperform their rivals (by whichever means) with the goal to maximise profits in the long run. However, traditionally public housing organisations, housing associations and cooperatives apply a non-profit model with the primacy of social goals and societal responsibilities. Hence, rather than vying for incompatible positions in the market, non-profit firms might actually be inclined or even legislated to behave cooperatively, with the result that the prospects and options of competitive behaviour are limited (see also Young et al., 2010). A further complication of the matter is that given the recent changes in the provision, funding, and management within most national housing systems in Europe (see, e.g. Mullins et al., 2012), many of these providers have endured a process of hybridisation towards becoming social enterprises, where commercial and social goals may conflict with each other. Recent studies (e.g. Blessing, 2015; Czischke et al., 2012; Teasdale, 2012) indeed show that the lines between non-profit and commercial activities across but also within these social housing enterprises are becoming blurred. What this implies for inter-landlord rivalry is part of the empirical investigation; however, one might assume that rivalrous perceptions and interactions are potentially stronger where the implementation of the social enterprise has progressed the furthest, both on a systemic and an organisational level.
Methodology and data
International case study city comparison
To guarantee for a large variety in landlords’ perceptions but also with the aim to identify how inter-landlord rivalry is framed by structural and institutional conditions, the paper is designed as an internationally comparative study. Here, the distinct systemic aspects of the rental housing systems in England and the Netherlands are used as two comparative cases. In addition, to account for diverging demand, supply and local policy conditions in which landlords operate, the study selected two (more or less) typical mid-sized local housing markets as case study cities for the empirical investigation: Coventry in England and Breda in the Netherlands.
The two countries and case study cities were primarily chosen because they form two distinct rental systems (see Table 1, see also Lennartz et al., 2012) – in Kemeny’s terms they are two opposing examples of an integrated market (the Netherlands) and a dual market (England). More precisely, there are important differences with regard to who provides social housing and private rental products. Housing associations are the sole providers of social housing in the Netherlands. In England, both housing associations and local authorities rent out social dwellings, yet varying significantly across municipalities. Coventry is an example for a local market where all social dwellings are owned by housing associations, and where one of them is a large stock transfer association (Stephens et al., 2008). In the private rented sector (PRS), small-scale individual suppliers dominate in England/Coventry, whereas it is split into relatively similar shares between individuals and large-scale corporate investors in Breda. Furthermore, housing associations have increased their market shares in the commercial rental sector in the recent past, hence becoming a third group of landlords in the commercial renting business. It holds true for both case study cities/national markets that housing associations are by far the largest providers, where they have a continuously strong position in the Dutch case in terms of their market shares (> 80%). Mimicking the wider development in the UK (Kemp, 2011), local private landlords in Coventry have expanded their housing stock significantly over the past 15 years, challenging the market power of non-profit landlords in all market segments.
Housing market structure and policy elements in England/Coventry and the Netherlands/Breda (2010).
Notes: aEstimation based on 2006 data.
Statistical office Breda uses a different classification than WoON and does not have the ‘other’ category.
In terms of the similarity between social and private rental products and the regulatory approaches to them, the two cases show distinct patterns as well. 2 Social landlords in England operate in a highly regulated environment with regard to how rents are set, which property rights tenants have, how tenants gain access to available dwellings and how much of their rent is covered by housing benefit. In contrast private landlords’ lettings are fully based on market mechanisms, where housing allowances are generous in general, 3 yet more limited than housing benefit payments. In contrast, the dwelling itself is the subject of rent setting and property rights in the Netherlands. Here a distinction is made between a heavily regulated segment and a deregulated segment, meaning that all landlords who operate in this segment are subject to the same regulatory rules, with the exception that a waiting list system applies in the social sector, 4 while willingness and ability to pay define the access to all dwellings owned by private landlords. The availability of housing allowances follows this dichotomous market structure (see also Lennartz, 2013).
To conclude, the brief depiction of the structural and regulatory conditions of social and private renting in Coventry/England and Breda/Netherlands suggests that inter-landlord rivalry should be stronger in the Dutch case because of higher product similarity and the existence of large individual suppliers in both rental sectors. On the other hand, the increasing market domain overlap of social and private landlords in the past two decades suggests a higher potential for rivalrous perceptions and interactions in the English case study. It is up to the exploratory investigation to unfold whether these structural aspects reverberate with local landlords and whether one of these aspects has a seemingly stronger impact on rivalrous perceptions and interactions.
Interviewing and data analysis
To explore landlords’ differing views on competition and competitors and the consequent competitive behaviours in the two markets a qualitative research approach was chosen. The empirical strategy builds on a series of semi-structured in-depth interviews with 36 social and private landlords. 5 Here, a purposive sampling strategy was applied (see Ritchie and Lewis, 2003): In the social housing sector, all housing associations that own and manage housing stock in Coventry and Breda were invited to join the study. Letters were sent (followed by telephone calls) to the respective head offices explaining the goals of the research and asking to be put in contact with managers working in the lettings, project development, business strategy and commercial units departments. Interviewing different research units (i.e. managers) within the same corporation aimed to get a maximum variety of perceptions and to account for different cognitive capacities. This sampling strategy resulted in the participation of eight housing association managers from all three major housing associations in Breda, while in Coventry eight managers from only the three largest housing associations were recruited for the interviews. 6
To cater for different market segments and business models in the PRS, the sampling in Coventry targeted side-line as well as business investors (see Crook et al., 2012). Seven of the nine interviewees were recruited through the private landlord accreditation system of the Private Sector Housing Department of the Coventry Council, while four of the nine participants were business landlords with mid-sized portfolios (>10 dwellings). In the Dutch case, the purposive sampling targeted small-scale individual landlords (three cases), corporate investors (two cases), as well as letting agents who cater for both landlord groups. The actual recruiting was achieved by cold-calling private landlords who owned at least ten dwellings in the city. 7 In both cases, the sampling targeted property managers who make day-to-day decisions on rent increases, investments, tenant choice, rather than those who merely own the property but outsource management decisions. As a result, two out of seven interviewees in Breda and two out of nine participants in Coventry were letting agents who took over the property management from local landlords. All of our participants will be called landlords, which implies that they could be managing the property without owning it.
In all interviews an open question topic guide was used that revolved around five key themes: products and tenants, perceptions of the other rental sector and other landlords, meanings of competition, goals and tenancy strategies, views on national and local housing policy. The first step in the analysis of the interviews was the full transcription, sorting and consequent indexing (coding) of all interviews with ATLAS.ti 6.2. Some examples of the codes that were applied are: ‘perceived competitive pressures’, ‘rent setting strategy’, ‘competitors in the market’ and ‘commercial business operations’. Through the help of memos and analytic notes these codes were then reorganised, linked with each other, and, finally, interpreted in the context of the different forms of rivalrous perceptions and interactions; i.e. the question of whether we can observe a relational, direct, subjective and non-reciprocal competitive behaviour.
Results
Perceptions of rivalry and rivalrous interactions
The general view of housing association managers in Coventry is that the PRS consists of various diverse market segments. They mainly refer to three submarkets: The student lettings market is seen as private landlords’ major field of activity, which spatially concentrates around the vicinity of university facilities. The market for professional households is perceived as a high-turnover market that serves customers who are more affluent but are not (yet) seeking to buy their own property. Finally, the bottom end of the market is considered to serve low-income households only. A recurring theme here is that at the lower end of the market the type of dwellings and the tenants that private landlords cater for are similar to social housing; yet, the property rights and price/quality relations of these dwellings are considered to be better than in the low-income PRS.
These views on market segmentation and product similarity generally reflect social landlords’ prevailing perceptions of who their competitors are. Housing association managers agree that there is no competitive relation with student housing and the more expensive professional housing landlords in the city. They acknowledge that there would be some scope for a competitive relation; however, owing to the fact that they are active in the market for professional households only on a very small scale, if at all, the notion of rivalry with private landlords in this segment plays a more or less insignificant role for them.
Contrarily, perceptions of rivalry with private landlords who operate in the low-income renting segment do exist. Yet, this was not the case for all association managers, where a noticeable distinction is that day-to-day tenancy managers (four interviewees) tend to see private landlords in the lower market segment as direct competitors, whereas general managers tend to observe no such relationship. We would argue that the varying perceptions within the same organisations provide some evidence for the thesis that the bias in the knowledge and information that managers obtain through their everyday work supports the theory of diverging cognitive capacities to understand and evaluate inter-landlord rivalry. Where perceptions of rivalry exist, they are largely based on incessant experiences with social tenants leaving the housing association for accommodation in the PRS, where the additional search, repair and vacancy cost are seen as the main problems of loosing these tenants.
Considering the strategic behaviour of social landlords, the interviewees state that they find it difficult to react to increased competitive pressures. Simply lowering rents or offering better dwellings is not an option that can be pursued. Instead, they have come to apply what could be called a ‘negative marketing campaign’, in which they actively try to convince tenants that they would lose out on property right and repair services when renting with a private landlord. Besides lacking specific possibilities to engage in targeted competitive behaviour, social landlords have no shared understanding of what it might be based on. Although they tend to agree that it does not take place on flexible rent setting and locational choices, there are different accounts of competition on waiting time, the availability of dwellings, quality, repair services or service charges. Again, this might signify how different cognitive capacities can lead to diverging ways of deciphering the market environment. Finally, social landlords only have a vague picture of whom they are actually competing against. Private landlords are regarded as an anonymous and unidentifiable mass of market actors, on whom little information is available and with whom they have only occasional or no personal contacts. Competitive behaviour is in contrast to the theory of inter-firm rivalry thus untargeted and mainly implies unspecific strategic measures.
We are seeing increasing competition from the private rented market, particularly over the last five years where there has been a shift to a lot of individuals buying properties and then let privately. Also, we can see that our rents have gone up, while their rents have come down, so we are closer together these days. (Coventry, social landlord)
Private landlords’ perceptions of social housing in Coventry are mainly steered by negative stereotypes and stigmatisation of both social landlords and tenants. Landlords operating in all market segments share a general suspicion of neglected dwellings, poor repair services and a waste of public money in the social housing industry. The views on social tenants are equally negative as there is a general expectation of antisocial behaviour and non-payment attitudes. More specifically, low-income-sector landlords are aware that they provide housing in similar locations for similar types of households, leading to a higher likelihood of inter-tenure moves of their own tenants. However, a common theme is that private landlords picture low-income households to prefer a social dwelling to their own accommodation if it was available, mainly because rent levels are lower and security of tenure is stronger for similar types of accommodation. It is thus unsurprising that all interviewed landlords, including low-income housing landlords, contradict the idea of inter-landlord rivalry. Arguably, this testifies to the non-non-reciprocal nature of rivalry, where competitive actions of social landlords do not provoke a reaction by their profit-oriented private counterparts. Notably, one might expect that a potential strategy of binding tenants to the sector is the lowering of rents. However, quite the opposite is true, as there are examples of private landlords who use a fixed rent policy in which rents are set to the maximum amount of housing allowance and tenant selection is based on referrals from the local authority, thereby keeping vacancy rates low and short. 8
Interviewer: You just described that there is a competitive relationship with other private landlords. Do you see a similar relationship with social landlords in the city? Landlord: No, no, they can get as many tenants as they want. My tenants would go to Whitefriars [the largest social landlord] if they could get a place […]. Interviewer: Is that a problem for you, people leaving to the social housing sector? Landlord: It hasn’t been when it happened. People move on. It is an accepted problem. If they pay £500 for my dwelling and can have a Whitefriars’ dwelling for £250, it is a no brainer. (Coventry, private landlord)
Similar to the situation in Coventry, social landlords in Breda do not see private renting as one coherent sector but as a set of different market segments, which are served by different types of providers. First, their categorisation of the PRS reflects the official distinction between regulated and deregulated dwellings. The prevailing perception here is that the largest share of all private renting is to be found in the deregulated high rent market segment, whereas the provision of low-income rentals is only marginal, with the student lettings market as the main exception. Remarkably, social landlords completely ignore the provision of regulated dwellings by institutional investors, which in reality amounts to several hundred dwellings in the city and in terms of product characteristics is very similar to social rentals. This is rather revealing for the self-perception of social landlords as the singular provider of affordable, regulated rental housing. Consequently, all interview participants do not see any competitive relationship with private landlords for lower-income households. On the contrary, in business practice one association in Breda goes as far as referring people who are stuck on the social housing waiting list to private providers. This service covers both the regulated and deregulated market segments and clearly shows that this specific association sees the private sector as complementary rather than in competition with its own provision. This behaviour signifies a tacit form of cooperation between social and private landlords and might be interpreted as the opposite of direct and relational rivalry.
On the other hand, strong perceptions of rivalry with private landlords do exist in other market segments. First, private landlords who provide student housing – which is in terms of price levels is often part of the regulated sector, but is targeted at the student population only – are seen as direct competitors. Even though housing associations only have periodical vacancies in their student halls, they see themselves in a weak competitive position, since private landlords tend to provide more desirable products – i.e. rooms in shared houses in attractive city-centre locations – and are therefore considered to be potentially better prepared to deal with declining demand. As a result, housing associations claim that there is little scope for reacting to the actions of private landlords because of the limited possibilities to change the product itself (e.g. through lower rents). However, to guarantee low vacancy rates on a long-term basis, associations use a letting strategy that builds on close ties with and direct referrals from tertiary education organisations.
Second, perceptions of rivalry in the commercial market segment prevail where housing associations operate in this segment on a substantial scale. Two of the three housing associations have significantly enlarged their portfolio in the deregulated sector in the recent past. Here, managers share the perception that private landlords are their direct competitors, which particularly applies to large corporate investors. According to the interviewees, competing on the turf of private landlords takes place on purchasing building sites for new developments as well as on rent and quality levels. With the ultimate aim to minimise vacancy levels, housing associations pursue a strategy of aggressive price competition that is directly targeted at corporate investors, supplemented with a strategy of offering dwellings with comparably larger outside spaces and stronger customer inclusion in terms of interior design.
Letting agents and private landlords have never taken us seriously, but since all associations together will have 1500 deregulated dwellings, we are going to be a major factor for them. (Breda, social landlord)
In contrast to private landlords in the English case, private landlords in Breda appear well informed about the social housing sector. They make a distinction between the three housing associations, know where they are located and are knowledgeable about their recent development activities. PRS landlords’ views on the performance of social landlords in this segment is predominantly positive, emphasising the relatively high quality of social dwellings, while they also expose less stigmatisation of social tenants than their English counterparts. The perceptions on the commercial activities of housing associations are more ambivalent. There are no general resentments against commercial operations; however, they claim that associations have lost sight of their true mission which is the provision of social housing for low-income households.
Although private landlords who operate the regulated rental sector see housing associations in a dominant position, they do not regard them as direct competitors. Similar to the views of social landlords, all types of private landlords express the idea that the demand for more inexpensive rental dwellings is so large that finding tenants for their properties is not an issue. Accordingly, the prevailing notion is that their provision complements rather than challenges the provision of the three housing associations. In the student and the deregulated markets perceptions of rivalry certainly exist but they tend to vary between landlords, which again testifies to the subjective nature of rivalrous perceptions. Regarding their own product, most landlords of student housing see themselves in a favourable competitive position; nonetheless, considering the high output of new student units by housing associations, they predict a more difficult market environment in which enduring vacancies might occur. The interviewees did not express a firm view on how they might change their business strategy in that case, but they acknowledged that lowering rent levels as a competitive strategy might be a potential option.
The commercial rental activities of social landlords are widely regarded as a new form of competition in the deregulated sector. The fact that housing associations have entered the market with numerous relatively cheap dwellings is seen as a major problem. Housing associations are considered to deliver a strong product with modern facilities that might make corporate investors’ units from the 1970s and 1980s undesirable. Generally, the provision of commercial renting by housing associations is seen as unfair competition, where the non-profit business model of housing associations is seen as the key issue. In contrast to corporate investors they are not bound to specified profit margins, which enables them to keep rents comparably low, thereby putting pressure on private landlords’ rent setting. 9 The reaction to increased competitive pressures has primarily been the realisation of an exit strategy by corporate investors, particularly in the €700 to €900 rent segment. Indeed, these investors have massively disinvested in their dwellings in less popular neighbourhoods, and where possible they have sold them to other investors or tenants, which led to a decrease of their housing portfolio from 3600 in 2001 to 2750 dwellings in 2011.
Just assume we want to buy a dwelling in a location where we have this [more expensive] association stock. When we make a taxation of the rent that we can ask, their provision becomes a major factor. Let’s say our dwelling could make 900 Euros and the association asks 750 Euros for their dwelling, we have a serious problem as an investor. (Breda, institutional investor)
Table 2 summarises the findings of the interviews on landlords’ perception of rivalry and how landlords’ views relate to the taxonomy of inter-landlord rivalry. First of all, significant differences between the two rental markets can be observed. In Coventry, the interviews with housing association managers and private landlords in the low-income segment suggest the existence of subjective and non-reciprocal form of competitive behaviour with differing perception within and across both landlord groups. There is little evidence for direct and relational rivalrous interactions, since strategic choices are mostly not targeted at landlords of the other sector, but play out in a more implicit manner. In the Dutch case, the distinction between the regulated and deregulated segments is a key aspect of the interviews. On the one hand, the empirical investigation proposes the existence of tacit cooperation or at least mutual acceptance of one’s activities without retaliation in the regulated segment. Housing associations have significant market power, which limits the possibility for competitive actions by private landlords and lettings managers, who instead apply a strategy of niche market provision. On the other hand, there is a direct and relational, but also subjective form of inter-firm rivalry in the deregulated and the student letting market segments, which is marked by competitive price and quality setting.
Inter-landlord rivalry in Coventry and Breda.
We can conclude that the relationship between the two landlord groups does not correspond fully with the taxonomy of inter-firm rivalry. Although we have found various accounts for the cognitive peculiarities of competitive perceptions (i.e. it may be subjective and non-reciprocal), it has varying behavioural components in the two rental markets and is more implicit than an explicit strategy of landlords. In the following section we will take the analysis one step further and discuss possible explanations for why these deviances from the theoretical notion of rivalry might prevail, which seeks to reveal more clearly the structural and institutional imperatives of inter-landlord rivalry.
The drivers of and barriers to inter-landlord rivalry in mixed rental markets
A first observation that should be taken into account is that rivalrous interactions and perceptions strongly diverge across market segments and not only the type of landlords. One salient finding of the in-depth interviews in the Netherlands was that, under the condition that both landlord groups operate on a substantial scale, inter-landlord rivalry is stronger in the deregulated market segment. Hence, we argue that a meaningful distinction is between barriers to rivalry in the low-income segment and barriers to rivalry in the fully commercial rental segment. Moreover, the barriers to inter-landlord rivalry can be subdivided into three categories: structural, policy-related and business-related barriers (see Table 2).
Considering the structural barriers in the lower lower-income segment first, the city comparison suggests that probably the most crucial barrier to competitive interactions arises from the innate high demand for social housing. With 10,000 (Coventry) and 8000 (Breda) actively searching applicants for social housing, housing associations have no need to gain a competitive edge, while private landlords who seek to provide affordable rental housing are likely to have a continuous stream of new tenants. In reverse this means that the situation might look different in weak local rental markets (Bramley and Pawson, 2002), meaning that where the demand for affordable rental units is generally low, both types of landlords might actually have to vie for customers. However, the interviews propose at the same time that focusing on city-wide social housing lists as a measure of demand for affordable renting is problematic in that it ignores a possible divergence between the highly popular and the least desirable neighbourhoods in the same local markets (see also Bramley et al., 2008). Therefore, inter-landlord rivalry should be considered as spatially contentious and unevenly dispersed across urban areas.
Conversely, inter-landlord rivalry is severely constrained through the prevailing structures of affordable rental housing supply in both rental markets. For one thing, there seems to be a barrier for private profit-oriented landlords that relates to their vulnerabilities when dealing with what they perceive or actually are ‘problematic’ customers. Lacking the means for and experience in counteracting antisocial behaviour, they are most likely struck harder by rent arrears and, therefore, often refrain from operating in this riskier environment. Moreover, the empirical investigation suggests that the deconcentrated and small-scale supply structure in the makes it difficult for social housing to recognise individual providers as their competitors, making targeted moves, unlike in an oligopoly, inherently difficult.
Inter-landlord rivalry in the low-income segment is very much a political, regulatory construct as it is a market structural contingency. Four factors seem to limit competitive behaviour specifically. First, the relatively high rent differentials for similar dwellings, which imply a higher propensity for tenants stay in a social dwelling, even if they had the desire to move to another dwelling. To understand the implications for rivalrous behaviour such rent differentials need to be understood in the context of national rent setting rules and the availability of housing allowances for low-income households (see the section on ‘Methodology and data’ as well). Generally, regulatory mechanisms in the UK restrict social landlords in Coventry in their rent setting strategies and thus their abilities to react to competitive pressures from the PRS. This rent rigidity in the social housing sector then suggests that rivalrous reactions and interactions are more politically determined than the result of structural disparities. 10 The interviews in Breda equally propose that the Dutch rent regulation regime impedes competition between landlords; yet, the rationale is a different one. The dual system of rent regulation provides various incentives for private landlords to solely operate in the higher-revenue and less bureaucratic deregulated segment, implicating a decreasing market domain overlap and thus fewer possibilities for rivalrous interactions over time. From the tenants’ viewpoint, the rent gap also relates to the question of who in fact pays the rent bill. Where large rent differentials are fully covered by housing allowance payments, tenants have strong motives and increased chances for moving to better private rental accommodation, which in turn increases the need for housing associations to behave ‘tenant friendly’.
The second regulatory barrier relates to the quality setting mechanisms in the affordable sector. On the one hand, both social and private landlords in Breda emphasise that the relatively high quality of existing social housing impedes investments of private landlords. Even if they were generally interested to invest in the deregulated rental sector, they would have to match the high building standards of social dwellings, implying high and profit-decreasing investment needs. On the other hand, undercutting the quality standards by developing accommodation only in the lowest rent sphere is also no viable option. After all, the existing quality point-based rent regulation stipulates that minimum standards need to be implemented, and extremely high urban land prices require a minimum rent income to avoid loss-making developments. Social landlords in Coventry experience the reverse problem. They claim that because of the English social housing subsidy system they can only deliver highly standardised, minimally equipped houses when developing social housing. The political impetus to keep investments as cheap as possible often leads to a mismatch between social dwellings and the products of private landlords as well as tenants’ housing preferences.
Third, inter-landlord rivalry is limited through the dualism of the needs-based allocation in social housing and the ability-to-pay-based allocation in private. Housing association managers in Coventry note that the needs-based foundation of the choice-based letting scheme (CBL) particularly adds to limited strategic choices of reacting to competitive pressures. In practice, the CBL in Coventry stipulates that when a sitting household expresses the desire to move, housing associations are not allowed to offer a bigger dwelling in a better location to a household with no urgent housing need, often pushing them into the PRS.
Lastly, rivalry may be limited by the policies of local authorities towards the construction of low-rent dwellings. On the one hand, private landlords in Breda identify the problem that the local authority as the biggest owner of developable land tends to ask the maximum market price for building sites. This ties in with the view that the absence of direct and indirect subsidies (e.g. cheap land) for private landlords to develop low-income housing further limits the chances that they will (re-) enter this market segment. The local authority in Coventry, on the other hand, is considered to have a ‘preferred partner policy’ that favours non-profit housing associations as developers of new affordable housing. 11
Finally, the empirical research finds that, in addition to the structural and regulatory barriers, strong rivalry may be thwarted by the business strategies and goals of both landlord groups, where the ‘non-profit/for-profit cleavage’ seems to play an essential role. Even if both landlord groups provided similar products and served similar customers, which is true in specific submarkets in Coventry and Breda, the social objectives still act as the primary business of most housing associations, meaning that often they are not allowed or they are not willing to pursue rivalrous interactions. However, the interviews show that the line between social objectives and competitive behaviour is considered to be becoming increasingly fragile and it is interpreted and applied differently by individual housing associations (see also Mullins and Pawson, 2010)
Barriers to rivalry in commercial rental housing
Barriers to rivalry in the commercial sector generally have a different tenor. Since the provision of either landlord group is bound to identical regulatory conditions in both markets, meaning that social objectives are at least not stipulated and ensured by government supervision, rivalry becomes more or less a matter of whether non-profit housing associations actually want to provide commercial rental housing and who the dominant providers in the more expensive market segments are.
In business practice, social landlords in both case study cities see the current property slump as a major limitation to increasing their commercial renting operations. The development of market rental dwellings strongly aims at both continuous rental income streams and capital appreciation. In times of unstable or even negative house price developments, social landlords are increasingly doubtful whether the goal of capital appreciation can be realised at all. Social landlords in Coventry add to this the constraints in borrowing private capital for such projects because of more risk-averse lending behaviour by the banks. Yet, the housing market slump in the UK and the Netherlands at the time of interviews should be regarded as temporary and not innate barriers to expanding their commercial renting operations and thus stronger rivalrous interactions.
Second, statements from various social landlords in both cities indicate that their social business paradigm forms a barrier to new investment in commercial renting. The requirement for large capital investments in commercial developments might dry up the cash flows that are necessary for running and investing in the social housing activities. This also ties in with the notion that there are conflicting strategies of commercial housing operations. Since commercial renting requires long-term capital investments, many housing associations prefer the development of houses for sale. Furthermore, one can observe differences between traditional housing associations and those that run the former council stock in both cities. In practice, the financial resources of the former municipal housing companies are bound to paying off long-term debts and to burdening maintenance investments, putting them in a relatively weak competitive position in the commercial renting segment.
Finally, the empirical analysis (once again) suggests that a potential driver of inter-landlord rivalry is the similar size and market share of different types of landlords. Housing associations in Breda clearly target customers of institutional investors by developing similar, yet slightly better and considerably cheaper dwellings.
Conclusions
The purpose of the present study was to present an alternative account of the concept of competition between social and private landlords in contemporary European housing markets. While most existing studies in the housing and urban studies literature analyse competition between social and private renting on the basis of the distinct market environments in which landlords operate, this paper is the first study to examine the behavioural components of their competitive relationship. By putting forward and testing the concept of inter-landlord rivalry as a direct, relational, subjective and reciprocal form of behaviour, the study proposes a new concept for a more integral analysis of shifting rental housing supply structures, i.e. for rental markets where the traditional and once well-defined roles of different types of providers are becoming increasingly obsolete.
Based on the in-depth interviews with 36 social and private landlords in two case study cities in England (Coventry) and the Netherlands (Breda) the overall conclusion of this study is two-fold: First, the study found that where perceptions of rivalry exist they are highly subjective and not shared (i.e. non-reciprocal) among all landlords. The strategic behaviour that follows from these perceptions is equally diverse and is in reality more implicit than explicit, as it is not directly targeted at other landlords. This is particularly true in the lower-income market segment, which is characterised by a dominant behaviour of housing associations and some sort of subordination of private landlords. On the other hand, rivalrous behaviour does have a different connotation in the commercial, more expensive market segments. The examination of the Dutch case indeed shows that decisions on rents, quality and advertising here aim to get a competitive edge over landlords in the other sector.
The second conclusion is that perceptions of inter-landlord rivalry and the consequent business strategies depend on local market and regulatory circumstances, differ across landlord types, and may be determined by the position of individual managers in an organisation. More precisely, in line with the original theory, the paper provides evidence that a dominant market position of few and large providers, low levels of social and private rental product similarity, and little market domain overlap are the main barriers to a rivalrous inter-landlord relationship – or in other words, it is significantly limited in markets that are signified by monopolistic competition structures. In reverse, there is evidence that rivalrous perceptions and interactions between different types of landlords are relatively stronger in more oligopolistic market setups – admittedly, this finding requires further testing to see if it holds in a wider set of market structures; such an approach would ideally consider and compare local rental markets in which both sectors are signified by a more deconcentrated supply structure than is the case in our two cities. Furthermore, it was shown that stronger regulation on rent and quality setting, entrenched relations of housing association managers with local authorities, as well as tenant selection and tenancy management limit the possibilities to pursue competitive strategies. Finally, the study points towards a potential adverse effect of the different business models of social and private landlords on inter-landlord rivalry. In principle, the non-profit business model of social landlords can be interpreted as a self-imposed constraint to competitive behaviour. Yet, at the same time the empirical investigation reveals that the hybrid status of housing associations may imply different approaches in mixed rental markets within the social housing sector: Housing associations that have kept a more (social) mission-oriented approach reveal little competitive awareness in mixed rental markets. On the contrary, associations with a more profit-oriented, commercial profile actively navigate the mixed market by applying competitive strategies where possible and cooperative ones where necessary.
As such, the study has potential implications for policy-makers and housing policy scholars who regard inter-landlord rivalry in mixed markets as a potential option for redesigning rental housing systems. First, the study suggests that to increase competition at the lower end of the rental market, one might reconsider: the relatively high financial risks for small-scale landlords and institutional investors when developing and buying properties in the affordable segment; the lack of public support in dealing with more ‘difficult’ tenants; the lack of investment opportunities in the existing built environment; the division of property rights and housing allowances for tenants that clearly steer demand towards housing associations; and – which is the case mainly in the Netherlands – the rent regulation system that distinguishes between a deregulated and a regulated segment, giving private parties little incentive to invest in the former. Also, if public policy aimed to increase rivalrous interactions rather than creating more competitive rental markets in general, the research has shown that policy-makers should discuss how to attract large-scale corporate investors to enter all segments of the rental market.
What has not been touched upon in this paper though is the question of whether more competition between the two landlord groups is desirable after all. 12 Readers will object – and rightly so – that more information is needed on the (unintended) negative societal consequences of a mixed market approach. After all, if the question of how to push private landlords (and possibly other social landlords) out of the market becomes the main concern of non-profit landlords, a complete demise of their social goals and obligations seems likely. Similarly, it has not become evident yet that profit-seeking actors are well equipped or even interested in providing rental accommodation with a social purpose. Hence, to clarify these issues, future studies should carve out whether a welfare-optimal balance between efficient and equitable rental housing provision exists, and if so, where this might lie.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.