comparing Denmark and Sweden from the
perspective of path dependence and change
Bo Bengtsson & Lotte Jensen
Path dependence is strong in housing institutions and policy. In both Denmark and Sweden, the universal and `unitary' (Kemeny 1995) housing regimes can be traced back to institutions that were introduced in the post- WW1 period and unfolded and consolidated after WW2. Recently, the universal and unitary housing systems in Scandinavia and elsewhere have been challenged by economic cutbacks, internationalization and ideologies of privatization.
Although the Danish and Swedish housing systems are both universal and unitary in character, they dier considerably in institutional detail. Both systems have corporatist features; however, in Denmark, public housing is based on associational tenant democracy and institutionalized bottom-up control, whereas in Sweden, it is based on companies that are owned by the municipalities, combined with a centralized system of rent negotiations. This chapter analyses the present challenges to the Danish and Swedish housing regimes and compares the responses and outcomes in terms of policy change and/or insti-tutional continuity (`path dependence'). Overall, the more decentralized Danish housing regime so far seems to has resisted pressures for change and retrenchment better than the more centralized Swedish regime have.
The chapter is based on a comparative project on the housing regimes of Denmark, Sweden, Norway, Finland and Iceland in the 20th century and recent research into con-temporary Danish and Swedish social and public housing governance (Bengtsson et al. [2006] 2013; for an overview in English, see Bengtsson & Ruonavaara 2010; Jensen 2011; 2013b).
A historical institutionalist perspective on housing policy
During recent decades, path dependence perspectives have been given growing attention in historically oriented social studies. Path dependence is often viewed as the basic causal mechanism in historical versions of institutional theory (e.g., David 1985; Hall & Taylor 1996; Thelen 1999; Mahoney 2000; Pierson 2004; Sewell 2005). The general idea is that if, at a certain point in time, the development takes one direction instead of another, otherwise feasible alternative paths will be closed or at least increasingly more
di cult to reach at a later historical point. Thus, inertia (institutional, social, economic and technological) is the main interest of path dependence analysis. In this chapter, housing policy is viewed in light of weak path dependence, i.e., a `historical pattern where one event, which is more or less contingent, considerably changes the probability of subsequent alternative events or outcomes' (Bengtsson & Ruonavaara 2010). This weak concept of path dependence situates the analysis somewhere between structural determinism (Mahoney 2000) and a loose claim that `history matters' (cf. Thelen 1999; Pierson 2004; Crouch & Farrell 2004).
In a weak version of historical institutionalism, `politics matters', in the sense that actors have perceptions, norms and strategies, mobilize power and make decisive policy choices but do so in a context where already existing paths frame their incentives and opportunities and sometimes overrule them. This implies a soft lock-in mechanism based on three components: e ciency, legitimacy and power (Bengtsson & Ruonavaara 2010; cf. Jensen 2003). Consequently, a solution to a relatively small-scale problem can in-stall institutional patterns that, over time, develop into a consolidated path. This may be because the solution is perceived as being e cient or because it has gained strong normative support and legitimacy. Alternatively, nally, the solution may be backed by strong interests because it reinforces their power position. If institutions develop into e cient and legitimate problem-solving mechanisms backed by strong actors, then they de ne robust paths over time.
In such an actor-based historical analysis, the typical case of path dependence is where actors more or less deliberately design institutions at point (or points) A, institutions which at a later point B set the rules of the political game between the same or other actors. In retrospect, the historical development can be perceived as an on-going and self-reinforcing chain of games between actors, institutional change, new games and new institutions.
As developed elsewhere (e.g., Bengtsson & Ruonavaara 2011; 2017), there are at least four arguments why path dependence would be stronger in housing provision than in other welfare state sectors. First, housing has some speci c features as a consumption and investment good. Houses and dwellings last for a long period of time, and they are tied to a speci c place, slow to produce, expensive, and not easily substituted with other goods (Stahl 1985; Arnott 1987). Second, the social importance of dwelling and the emotional, social and cultural attachment costs of moving have a stabilizing eect on the market (Dynarski 1986). Norms of social exclusion and rules of eligibility also work as obstacles to change. Third, market contracts serve as the main mechanism for distributing housing, whereas state intervention typically takes the form of correctives, de ning the economic and institutional setting of these contracts (Bengtsson 2001; cf. Torgersen 1987, 116118). Hence, the main political institutions in housing are tenure forms and other types of regulations that de ne the basic rights of possession and exchange, which is why some political self-restraint may be expected in a capitalist society. Fourth, the fact that housing is ultimately distributed in the market may in itself serve as a constraint to political change. To be successful, not only does a new tenure form need political support, but consumers must also be prepared to pay for it in the market and producers to supply it.
more of the underpinning mechanisms of e ciency, legitimacy and power are attacked frontally or undermined slowly. There are basically two routes to change: external and internal.
An external in uence, rst, implies direct `shocks' such as war, revolution, economic crisis or natural disaster. For example, WW2 had a severe impact on the housing shortage and subsequently on the political investments in housing policy after the peace. More recently, wars and natural catastrophes in other parts of the world have led to a consid-erable migration that can change the economic, social and cultural conditions of social housing areas in receiving countries. Recently, in many countries, economic crisis has severely impacted the ability of citizens to buy and rent and the price of investments and loans for construction.
Internal challenges emerge when the mechanisms on which the institutional paths are built become worn out by internal contradictions or increasing inability to provide answers to the problems that they were once installed to address or to new evolving forces or ideals. For example, economic subsidies installed to obtain broad tenant compositions can end up as a black market because tenants and landlords accept illegal money for housing, and governance forms installed to secure tenant in uence can stall due to a lack of motivation and capacity among tenants. As, e.g., Streeck and Thelen note, mechanisms of incremental change such as `layering', `drift', `conversion' and `exhaustion' can also have transformational consequences over time in a generally path-dependent context (Streeck & Thelen 2005).
The disintegration of the Norwegian universal housing regime, based on price-regulated cooperative housing, is an interesting example of what may occur with worn-out housing institutions. Annaniassen (2013) claims that this system was doomed from the very beginning since it unintentionally produced antagonistic forces, namely, a large number of cooperative dwellers who tended to de ne themselves as owners, which eventually made it politically impossible to retain the price control (cf. also Sørvoll & Bengtsson 2018).
In the following, we discuss the Danish and Swedish unitary housing regimes and their contemporary challenges from this historical institutionalist perspective.
The Danish housing regime social housing and tenant participation
For a long period of time, Danish social housing has been based on independent, but pub-licly subsidized, collectively owned housing associations with a strong tradition of internal tenant democracy at all levels and a cost-dependent rent level for each housing estate, unattached to the local market of private rentals. There are restrictive legal limits to activities in addition to construction, renewal and management of the housing stock and rm demands on the reinvestment of means in new housing or refurbishment. The target group is broad, on top of the commitment to serve as a policy tool for the municipality to assist socially vulnerable groups, the elderly and the handicapped in return for public subsidy.
Denmark introduced loans for municipal housing by the end of the 19th century, but
due to local government resistance, the impact was weak. In this situation, the labour movement changed its strategy to support for social housing associations, initially
orga-nized by labour unions and charities, based on the `self-help' principle. However, Den-mark had early experiences with speculative sales of entrepreneurial associations that had achieved public support. Hence, as early as 1919, legislation was prepared to draw a demarcation line between housing associations that were worthy and not worthy of subsidies. A condition for public loans was that the public money should stay in the as-sociation. Hence, tenants were ascribed a stakeholder-user role in the association, de ned as the collective owner, consequently without individual rights to capitalize any value increase of the at through sales. Once the mortgage is paid o for older estates, rents are channelled to the National Building Foundation (NBF), which was established in the 1960s and funds renovations and social programmes in social housing estates. Subsidies and regulation have changed format across history but remain in place in 2016.
Early on, the single housing estate within the associations was de ned as a separate unit with its own economy. Subsequently, the physical estate units became political units as well, as tenant democracy was developed in the 1970s. At present (2016), approxi-mately 530 non-pro t social housing associations exist, representing approxiapproxi-mately 7,000 democratically elected local boards related to the physical housing estates and legally and economically independent from other estates in the association. Both estate-level and association-level entities mobilize a large number of tenants in local matters and housing policy at large (Jensen 2011). Social housing organizations host approximately 20 percent of the population. The bottom-up construction of the system implies the in-clusion of tenants in management rather than the mobilization of tenants independently of management.
The Danish unitary housing regime can be summarized as follows:
Collectively owned, independent, but publicly subsidized and regulated housing associations.
Universal access to subsidized housing via waiting lists, unless special agreements apply to ensure a balanced composition of household types. Municipal access to a minimum of 25 percent of vacancies.
Signi cant corporatist in uence on housing policy via the housing associations' um-brella organization BL.
Institutionalized tenant democracy at all levels of housing associations but relatively weak independent tenant organization outside the institutionalized system.
The Swedish housing regime municipal ownership and corporatism
The Swedish housing regime is universally oriented, without individual means testing. Municipally owned housing companies managed at arm's length from political in uence are the mainstay of the system. There are formal links between rent-setting in the public and the private rental sectors within a `corporatist' system of centralized rental negoti-ations between landlords (public and private) and representatives of the strong national tenant movement.
Swedish tenants have been strongly organized since the period between the two world wars, as the deregulation of rents led to signi cant rent increases in a period of unem-ployment and housing crisis. This mobilization impacted on the formation of Swedish housing policy for the rest of the century and onwards. The institutionalized corporatist position of the national Swedish Union of Tenants (SUT) is based on not only membership which is uniquely high but also the broad acceptance of the organization as a liable counterpart that can be expected to look more widely than to short-term member and organizational interests.
The central role of municipal housing companies (MHCs) dates back to early post-war legislation where originally, as in Norway, housing cooperatives were also given an important role. Unlike Norway, however, Sweden prioritized municipally controlled hous-ing. Municipalities were not legally obliged to establish MHCs, but de facto, all large municipalities did (Bengtsson 2013, 114116).
Arguably, the most de ning aspect of the Swedish housing regime is its consumer-based corporatism. Virtually all rents in Sweden are set in collective negotiations between landlords and tenant unions. The legal mainstay of this order is the use value system of rent setting (introduced in 1968), where a claimed rent can be rejected as unreasonable if the regional Rent Tribunal (where the organizations are also represented) nds it to be considerably higher than the rents charged for dwellings with the same use value in the same local market.
Following the new EU-inspired legislation of 2011, the Swedish MHCs and the rent-setting system are at a crossroads. Previously, MHC rent negotiations were based on self-costs and negotiations on private rents on use-value comparison with MHC rent levels (see Bengtsson 1994 for a presentation of the logic behind the system). The new legislation, however, prescribes that MHCs are to act on `business-like' principles, and use-value comparisons can be made with both private and private rents provided that they are set in collective negotiations.
Thus, the new legislation challenges the role of MHCs as price leaders in the local rental markets. The use-value system has given MHCs a stronger impact on the rental market than that indicated by their market share of ca. 20 percent of the housing stock approximately the same size as the private rental sector. (In both Denmark and Sweden, owner-occupation is the largest tenure form, with between 50 and 60 percent of the housing stock.) The further political implications of the new legislation are discussed below.
The Swedish unitary housing regime can be summarized as follows:
Universal access to publicly subsidized housing. The general housing subsidies have been successively phased out since the 1990s, but there is still no means-tested sector aimed at low-income households.
Municipally owned housing companies, previously cost-based but now more `business-like'.
An integrated rental market, with the use-value system as the formal link between public and private rent-setting.
Corporatist in uence on rent levels and on housing politics based on a strong tenant movement. Only weak local tenant participation.
Table 1. The dierences between the Danish and Swedish housing regimes
Formal Municipal Corporatist Tenant in uence
organization in uence in uence options
Denmark association some means-tested strong via umbrella strong direct local dwellings for social organization BL in uence option qua
purposes tenant democracy
Sweden MHC (LLC) ownership of MHCs; no strong via SUT and weak and indirect institutionalized means- centralized rent local in uence
testing negotiations
Comparing the Danish and Swedish housing regimes
The dierences between the Danish and Swedish housing regimes are summarized in table 1.
The dierence in organizational form between the two housing regimes stands out. The Danish regime is built on associations of tenants with a strong right to in uence local and associational matters in corporatist housing organizations. The Danish municipalities can utilize housing associations as their policy tool when helping vulnerable groups, but they have no ownership in uence. This is in contrast to Sweden, where municipal ownership is balanced by a corporatist in uence from a centralized tenant movement. In condensed form, independent associations are the power core in the Danish public housing regime and a duality of municipalities and tenant organizations in the Swedish regime. Below, we discuss how these dierences have played out in the contemporary period of pressure on the unitary model.
The history of housing policy follows a parallel, even if not simultaneous, development curve in both countries and clearly in many others, including the Nordic neighbours of Norway, Finland and Iceland (cf. Bengtsson et al [2006] 2013). The introduction phase in the beginning of the 20th century, where housing policy emerged as a public
concern and became politicized, was followed by a construction phase during the decades after WW2, when large-scale political programmes were launched around the provision of as many housing units as possible at low prices and the national housing regimes were institutionalized. After the 1970s, the regimes entered a management phase, and political attention was directed towards management of the already built environment and the social eects of the physical structures, followed by a retrenchment phase, in which the post-war institutions came under considerable political pressure.
The unitary model institutional characteristics and political vision
Although the Danish and the Swedish housing regimes are very dierent, they are both often de ned as `unitary', following Kemeny's (1995, 4) ideal-type distinction between
`unitary' and `dualistic'. According to Elsinga, Haner & van der Heijden´s (2008) elabo-ration of Kemeny's model, the unitary model is characterized by (1) competition between the private (pro t) and the public (non-pro t) rental sectors; (2) the provision of rental housing to broad layers of the population; (3) market-dependent rent levels and rent dif-ferentiation in the non-pro t sector at a level that is lower than the market rent; and (4) limited market segmentation. In this model, the private and public sectors compete over attracting all types of households, including more well-o strata of the population. Until 2011, the Swedish rent setting system represented an unusually clear cut even formalized case of a `unitary social rental market' according to Kemeny's model. The Danish regime also has most of the elements of the unitary model, although it includes a moderate element of means-testing. Regarding Elsinga, Haner and van der Heijden´s third criterion, public rents may actually in some areas be higher than private rents in both Denmark and Sweden.
In contrast, the dualistic model separates non-pro t from market rental housing. There is no competition between the two sectors, and the non-pro t sector is regarded as a social safety net. The rent level in social housing is not market-dependent, and the demand is by de nition higher than the supply, which necessitates dierent forms of administrative allocation via queuing and point systems. The rental market is typically strongly segmented, and social housing is allocated to households that are unable to nd acceptable housing in the market (Elsinga, Haner & van der Heijden 2008; cf. Stephens 2016; 2017 for a critical discussion of the relevance of Kemeny's typology today).
The vision behind the unitary model is typically claimed to be one of using housing policy as a means of societal integration between dierent groups of citizens by providing good and equal living conditions. In both Denmark and Sweden, this was an explicit goal when the dualist policy of providing special housing units for the elderly and for poor people with many children was explicitly abandoned after WW2 (Indenrigsministeriet 1945; Boligministeriet 1987; Bengtsson 2013, 114). Hence, the unitary model goes beyond the idea of distributional equality and builds on broader integrative ideas of housing policy than merely providing shelter. In modern times, this is mirrored, for example, by the European Federation of Public, Cooperative and Social Housing (CECODHAS) in its response to a debate over state aid and EU competition rules: `For us to deliver sustainable neighbourhoods, a long-term perspective and local development strategies designed together with residents is needed. [. . . ] To provide a suitable response to the most vulnerable, giving them the same opportunities, we need, and want, to do more than provide a roof for those in need. In that context, imposing very low income ceilings as an eligibility criteria for accessing social housing constitutes a counterproductive approach and will only help to create the ghettos of tomorrow' (CECODHAS 2011).
The contemporary challenge to the unitary housing regimes
A historical institutionalist account of contemporary Danish and Swedish housing policy delves into the mechanisms of and responses to retrenchment i.e., the political and economic challenges to the core features of the unitary housing regimes.
Sweden experienced pressure on the price gap between homes with dierent historical -nancing conditions. Consistently, older apartments became more attractive options than newer apartments, regularly due to a mix between price (historical costs and rent regula-tion), location (inner city) and building style (contrast to suburban concrete).
The Swedish use-value system made it possible to level out the cost dierences to some extent through the transfer of capital from more to less pro table estates within the same MHC (Abbas 2002; Bengtsson 2013; Elsinga & Lind 2013). Denmark attempted to equalize the costs of living in older and newer apartments by introducing the NBF in 1966, where older associations pay a compensational `fee' to a common pool that is spent on co- nancing new associational construction and the refurbishment of older and worn-down stock (Jørgensen 1994, 190). Although rent gaps continued to exist, this model created a closed money circuit with the intention of increasing the self- nancing component of the social housing sector.
However, in the retrenchment phase of the most recent decades, the two housing regimes faced even stronger challenges. Although the problems of economic e ciency and equity between housing of dierent ages, tenures and locations remained, the po-litical discourse now took a neoliberal turn where all types of market correctives and interventions were viewed with suspicion. Additionally, internationalization, particularly Europeanization under the EU regimes, has recently played an increasing role in national housing provision. Additionally, increasing migration has heavily impacted the availabil-ity of social housing and the political housing discourse.
The retrenchment of the Swedish housing regime
The Swedish retrenchment phase has come in two waves (to date), one in the early 1990s and the other 15 years later. In 1991, the incoming right-wing Bildt Government launched a `system shift' in housing policy, acknowledged by the closing down of the Ministry of Housing and featuring the dismantlement of state housing loans and universal production subsidies. This reform had urgent economic motives, but it was also based on the neoliberal assumption that competition and freedom of choice were the precondition for all citizens to obtain a good home at a reasonable price. These liberalization measures were not reversed with the return to o ce of the Social Democrats in 1994.
The corporatist rent-setting system was also attacked by the Bildt Government, but it was sustained after a successful campaign launched by SUT and backed with opinion polls that indicated that the majority of Swedish tenants were against the proposed changes and `market rents' more generally. Eventually, only minor adjustments were decided on, formally weakening the monopoly position of the tenant unions, but in reality without much eect (Bengtsson 2012, 10).
After 12 new years in o ce, the Social Democrats were defeated in 2006 by a right-centre alliance led by the Conservatives. Now, subsidies to housing construction were -nally phased out, and new legislation was introduced that made owner-occupation in new-built multi-family housing possible but still not the conversion to owner-occupation of existing rental estates. Instead, conversion from rentals to cooperative tenant-ownership was supported, and, in particular in Stockholm, a large number of MHC estates were
transformed to market-priced cooperatives. Again, however, the corporatist rent-regulation regime survived, and the right-wing Government did not propose a more dualistic model of means-tested social housing (Bengtsson 2012, 12).
Thus, the Swedish corporatist housing regime was still and, to some extent success-fully, resisting marketization when the European Union's competition policy and legisla-tion against state aid, the so-called SGEI legislalegisla-tion, entered the nalegisla-tional agenda. What set the ball rolling was when the Swedish Property Owners, the umbrella organization of the private estate owners and landlords, via the European Property Federation, formally reported the state of Sweden to the European Commission for giving state aid to the MHCs, contrary to EC law. The argument was that the MHCs received economic sup-port from the municipalities, although they competed in the same market as the private landlords. Given that public subsidies had already been largely dismantled, the com-plaint was based on an interpretation of state aid as a case where a municipal owner of an MHC `. . . did not demand a market-based rate of return on the market value of the asset of the company' (Elsinga & Lind 2013). Due to use-value-based comparison, private owners were unable to capitalize, e.g., on the potential market value of inner city ats in Stockholm. By de nition, then, MHCs were subsidized if they did not charge rents that led to maximum pro ts.
Subsequently, in a voluminous report from 2008, a government committee on the rela-tionship between Swedish law and the SGEI legislation concluded that the rent negotiation system was problematic since MHC rents aected private rents. Two alternatives were presented, `business-oriented MHCs' and `cost-oriented MHCs'. In the former, MHCs were to be run according to business principles, much like private property owners. They would only be allowed to take policy-motivated measures if these were pro table in eco-nomic terms, at least in the long run. The `cost-oriented' alternative would be feasible only if Sweden could be exempted from the general EU ban on state aid, based on a speci c assignment to provide housing for all, to promote housing integration and to oer tenants in uence on their housing and MHC landlord (i.e., a unitary rental market logic; however, this alternative was not actually tested). In both alternatives, MHC rents would no longer serve as the norm in use-value comparisons but be replaced by negotiated rents in general. This emphasis on collectively negotiated rents again illustrates the general acceptance of the corporatist system of rent-setting. Needs-tested social housing should still be avoided (SOU 2008:38; cf. Bengtsson 2012).
In only three years' time, the o cial Swedish position and the Swedish housing dis-course had gone from viewing housing as a national concern to a somewhat anxious discussion about what would or would not be accepted in a potential trial based on SGEI legislation. A lengthy and complicated tactical process ensued, ending up with the SABO (the national organization of MHCs) and the SUT jointly declaring that they accepted the proposal of the Government committee that the use-value system should be kept but based on negotiated rents and not only MHC rents. Further, according to the joint declaration, MHCs should be run on business-like principles but still promote municipal housing provision. This 'historical compromise on rents' meant that the SUT had relin-quished its persistent defence of the self-cost principle and the priority of MHC rents in use-value comparisons. This opened up the locked political position; when the Property Owners also joined their old corporatist counterparts, the compromise was integrated in
a new Law on Municipal Housing Companies. This law, which has been in force since 2011, prescribes somewhat paradoxically that MHCs are to be run on business-oriented principles and, simultaneously, are to promote housing provision in the municipalities. Use-value comparisons are to be based on negotiated rents in general, not exclusively MHC rents.
The national organizations in the rental market had once again acted as the driving force in Swedish housing policy, this time in promoting a thorough-going change of the system that they built up and implemented over some 40 years. Importantly, however, even after the change, the organizations are still in control of rent-setting. In contrast, The Swedish Association of Local Authorities and Regions (SALAR), representing the municipalities and owners of the MHCs, was still left outside the three-party corporatism of the rental sector (see Pagrotsky 2010 for a discussion of the process based on retrospec-tive interviews with the involved actors). De facto, the new legislation, resulting from the `historical compromise', leaves ample room for local bargaining over the interpretation of `business-like' and `supporting municipal housing provision'. Additionally, to date, no politicians advocate a dualist housing regime with full-blown market rents and a residual housing policy based on means-testing.
After seven years with the new legislation, it is still far from certain what will be the new equilibrium of the Swedish rental sector. The implementation process is still on-going in the interaction between central actors, particularly in the collective rent-setting arena and the municipal housing provision arena. In the rent-setting arena, the collec-tive parties are currently testing each other in tough and sometimes failed negotiations, and in the housing provision arena, local norms and interpretations concerning what it means to `promote housing provision in the municipality' are being developed (cf. Bover-ket 2017). Bengtsson (2016) outlines three alternative ideal-type scenarios for the future development: resistance, adjustment, and system shift. The `resistance scenario' would largely imply the status quo, the `adjustment scenario` would mean a further weakening of the unitary logic, and the `system shift' would mean abandoning it completely. In prac-tice, the implementation so far has increased dierences in policy between municipalities (Grander 2015).
The Swedish case provides an example of the mobilization of political pressure by opponents to the philosophy underlying the universal welfare state and the unitary model and hence an attack on the legitimacy and power base of the existing system. The rst wave of criticism resulted in de-subsidization and privatization, notably in attractive inner city capital areas, whereas the second wave employed internationalization to adjust rent-setting to more closely approach market rents across sectors. This in turn led to a modi cation, but not a dismantlement, of the corporatist rent-setting system.
The retrenchment of the Danish housing regime
The Danish retrenchment phase seriously commenced in 2001, when a Liberal-Conservative minority Government took o ce after eight years of Social Democratic rule, with the sup-port of the nationalist welfare-oriented Danish People's Party, which had a considerable following among former Social Democrats and in social housing areas. As in the Swedish
case, two waves are discernible. The rst wave was almost an ideological tsunami, as the incoming Government rushed to ash privatization of social housing as one of its primary agships and point-blank refused to continue the historical path of incorporating organized interests in housing policy making: speci cally, the national organizations of housing associations and of local government were now ousted. The Ministry of Housing existing since 1947 and viewed by the incoming Government as an integral part of the So-cial Democratic project was thoroughly dismantled and its portfolio distributed across a number of other ministries. Instead, a `Ministry of Integration' was launched to signal the new Government's concern with ethnicity in the social housing sector, and all social/urban programmes were subsumed under the notion of `integration'. In summary, the policy was launched as an outspoken break with what was viewed as a Social Democratic housing regime. All tenants should be allowed the opportunity to buy their dwelling, either as individual or collective share-holder property, and become owners rather than parts of the political system of collective ownership within the sector, `from cradle to co n'. However, the policy immediately foundered on the, in constitutional terms, private status of social housing associations de ned as legal owners. Thus, legally, it soon turned out that the Government could not enforce privatization of real estate that was already private.
A longwinded political struggle between the Government and the social housing as-sociations and their national interest organization, BL, followed, including a spectacular political and media campaign based on the mobilization of the tenant hinterland i.e., 20 percent of the nation's households against the Government's plan B, in which individual housing estates democratically independent departments of the housing associations would be able to vote to put themselves on the market. Additionally, the national organi-zation of municipalities, which depends on the social housing associations as its primary social housing policy tool, mounted criticism and foresaw possible segregation eects (Jensen 2013a).
Hence, two narratives of housing policy stood against each other: a liberalization nar-rative, where limitations on the market and free individual choice were viewed as the problem, and a more Social Democratic narrative, which viewed the alleged consequences of the solution in the rst narrative as the problem itself: social segregation and the dismantlement of organizational neighbourhood structures. In turn, the political mobi-lization of tenants in uenced the Danish People's Party, keen to maintain its support among tenants, to soften up and support only a sunset legislation of `experimental' na-ture, allowing social housing estates to make decisions about voluntary sales within a de ned time frame. The expectations on the part of the Government were still signi cant and outspoken, but the sales policy never gained momentum. BL claimed that only the association level, as the legal owner of the property, and not the tenants of an individual estate, could decide to sell. Consequently, when the rst local housing estate made the decision to sell, the association led a court case that led to a protracted period of doubt about the future status of sales. Moreover, the Government had overestimated the resi-dents' interest in converting their tenant status into ownership, particularly because there were no signi cant economic gains attached and tenants would lose the service embedded in normal social housing management. By 2009, only a total of 62 dwellings of nearly 600,000 had been sold (Jensen 2012). Subsequently, the sale of social housing has become embedded in more strategic urban development schemes, subject to municipal plans and
decisions and joint acceptance between municipalities and associations rather than as a consequence of individual tenants wanting to become property owners.
The political implication was that the Government turned its attention from the pri-vatization of existing stock to limitations on future building policy. In 2003, a regulation of construction prices per square metre was issued, tying public subsidies from local government to a restrictive price ceiling, which de facto limited the ability to build. Sub-sequently, in 2005, the Government decided to signi cantly supplant public subsidies for new constructions with money from the NBF for the new construction of primarily housing for the elderly, i.e., expenditures that had normally been covered by the state budget. The social housing associations saw this as `milking' their capital savings and ex-acerbating their ability to invest in much needed housing refurbishment. In line with the sales policy, the social housing associations mobilized their political hinterland in a cam-paign pitched by the slogan 'government steals tenant money' (Nielsen 2010ba; 2010ab). Interestingly, Nielsen relates her analysis of the Danish development in 20012008 to Thelen and colleagues' ideas about incremental change in the forms of drift, layering and conversion.
In 201115, Social Democrats and Social Liberals were in o ce, and the Ministry of Housing was reinstalled, only to be discontinued after a swap back to a liberal minority Government in 2014. Demonstrably, a Ministry of Housing is consistently interpreted as a Social Democratic institution with close ties to the housing movement. Shortly after, the pressure on social housing took a new form, this time, through legislation on phys-ical planning. In the existing legislation, municipalities were entitled to de ne up to 25 percent of dwellings in a local area as being designated for social housing development. In February 2016, the coalition of Liberals, Conservatives and the Danish People's Party launched a proposal to abandon this clause. Immediately, the coalition of housing associ-ations and municipalities, accompanied by urban developer rms, mobilized to preserve the idea of integrated/mixed local environments (rather than radically segregated), ar-guably the backbone of the unitary model. In the end, the change was abolished in the legislative package.
With some time lag, the Danish case has repeated the Swedish experience, namely, ide-ological and political pressure from an incoming Liberal-Conservative Government wishing to privatize social housing associations by converting the role of the citizen from tenant to owner. The historically founded model of publicly subsidized but legally independent associations made it impossible to immediately force a direct privatization through, as the property was already formally private. As in Sweden, this exempli es a power struggle between adherents to and opponents of the unitary model. However, the Danish model proved di cult to break down, and when the Government turned to a plan B of indirect privatization and downgrading social housing construction, this led to vehement political counter-mobilization. Subsequently, a reform was passed, rede ning social housing associ-ations from objects of government steering to subjects in interactive governance processes aiming at creating sustainable neighbourhoods. The future strength of the social housing movements' role in a broad and unitary social housing policy will depend on its ability to avoid imploding the mechanism that maintains the power and legitimacy of the so-cial housing sector: tenant engagement through the internal democratic recruitment and governance system.
The open war over how the role and purpose of NBF should be interpreted paved the way for a preliminary time out, starting in 2006. The Government initiated a report on the future role of social housing, including vague ideas of copying the Dutch model of self- nancing social housing associations. Slowly, the Government started to wind down its exclusion of organized interests from politics and invited them into the writing process, realizing how much local social policy depends on the housing associations, not least because they began to claim ownership of the social role and responsibility as a key source of legitimacy. Representatives of the national and local government and of social housing could agree that the social housing sector should play a more `independent' or `self-sustaining' role, but in parallel to the Swedish case this Judgement of Solomon created considerable interpretational space in regard to the de nition of these key con-cepts. The Government could view it as an `end of subsidy' à la the Netherlands. The housing associations leaned somewhat towards the idea of enhancing the market position among the middle layers of the population through ancillary activities and more exible conditions for tenants, whereas the local government equated `independence' with less regulative state interference in local social housing policy development. The debates and the report writing bifurcated into two streams nance and governance of which the latter was considered easier to move forward without con ict. This in turn gained in-creasing importance for the future de nition of social housing policy. During the process, the BL's approach metamorphosed from a confrontational campaign organization, with strong links to the political opposition, into a more broadly de ned and more pragmatic interest organization, aiming to reach favourable agreements with the Government. Thus, by 2010, both price ceilings and the government use of fund money were modi ed although not removed.
The main conclusion of the eventually three voluminous reports was an interpreta-tion of `independence' leading to a new governance regime: the dialogue steering model, implying a soft version of `contract steering', in which the local government and the local housing associations reach agreement on housing challenges and mutually binding pol-icy measures in compulsory yearly meetings. In the new rulebook, housing associations are still the main tool of local housing policy, as a minimum of 25 percent of vacancies continue to be available for social purposes. However, the associations are simultane-ously re-de ned as co-players in local housing policy with the explicit task of countering segregation, providing housing at a reasonable price for all households in need of it and ensuring tenants' democratic in uence on their living conditions (Jensen 2011). (Similar assignments are often noted as examples of how Swedish MHCs should `promote munici-pal housing provision'; cf. Grander 2015.) Subsequently, the local governance processes, on the initiative from the Finance Ministry, have gradually been rede ned to monitor the cost pro les of social housing associations, based on `best case' logics that exert pressure on the more expensive associations to be able to cut on the housing allowances calculated as a function of the rent level.
In short, from being de ned as the core problem in 2001, the social housing associations were viewed as part of the solution in 2008 by the same Government. The Social Democratic Government in o ce from 2011, in light of the nancial crisis, decided to front-load investments in social housing refurbishment, in agreement with the new opposition. By summer 2012, the BL's CEO could celebrate `a good year for social housing policy'
with a consolidated role of social housing based on `a historically broad consensus about all important housing policy decisions, providing the safety and stability necessary for longsighted policy solutions' (DAB 2012). Although the new liberal minority Government started out by repeating the dismantlement of the Ministry of Housing, the political mobilization of the social stakeholders has given a hard time to liberalization strategies of curbing the future development of social housing through regulated prices, subsidies and, recently, access to urban space. The challenge for any government is that the social housing sector remains the key tool in the tool box of social housing policy.
Concluding discussion the jury is still out
Although the challenges to the unitary regimes have been similar in Denmark and Sweden, we can also observe important nuances. Although the problems of economic e ciency and equity have largely been the same, the Danish system was for some time strongly contested ideologically by the Liberal-Conservative Government. In contrast, the Swedish system has been and still is under pressure from the European SGEI legislation. In both cases, path-dependent resistance has been strong, although the more decentralized Danish housing regime overall seems to have resisted pressures for change and retrenchment better than the more centralized Swedish regime.
The Swedish housing regime came under political and ideological re in the 1990s, which led to declining subsidies and signi cant marketization pressure on inner city dwellings. In the early 2000s, property owners activated international mechanisms to enable pro table investments. Although the compatibility of the Swedish housing regime with the SGEI legislation was never actually tested, the voluntary adjustments initiated by the national organizations placed the rental market in a political limbo. On one hand, the corporatist model of in uence remains legitimate and eective, at least outside Stockholm. Additionally, local actors have considerable leverage for interpreting the new `business-like' approach of MHCs and de ning how it can be combined with sustained social responsibility. On the other hand, it remains to be seen how the Swedish unitary rental market will develop with no subsidies, increasing market conformity in rent setting, and municipalities less willing to cater to marginalized groups.
In Denmark, social housing policy moved from looking into the political abyss due to severe ideological attacks to re-entering a central policy position, continually coping with dierent types of pressure through internal and external mobilization. In contrast to Sweden, the Danish social housing sector, to date, has not been aected by the SGEI legislation. The housing regime has not been commented on by the Commission, and the common wisdom seems to be that the Danish construction of social housing does indeed meet the SGEI demands. On the one hand, restrictive limitations on additional commercial activities apply; thus, no criticism of cross subsidizing can be raised. On the other hand, the local government is guaranteed a minimum of 25 percent of vacant dwellings, which can be allocated to citizens with special needs. Regardless of formal applicability, there are no powerful stakeholders, comparable to the Swedish Property Owners, with a strong interest in the issue.
rather, the dependence on its power base: the active tenants who embody the central ner-vous system of political legitimacy and force. As marketization, in the shape of competi-tive private ownership dwellings, kicks in for the more resourceful social housing tenants, the political capital threatens to dry out. Thus, it may be the case that the social hous-ing sector has achieved an equal status as a local policy player but nevertheless stands to lose power if its constituencies weaken and to lose legitimacy if, in the absence of active tenants, professionals in the housing associations' management teams continue to take the seat as the real players.
Although the jury (and the Commission) is still out, it seems that the historical and path-dependent strength of the Swedish housing regime the direct public ownership and the unique and mobilizing bargaining system has become its Achilles heel under changing political conditions. In contrast, the Danish non-public/non-private model has allowed the social housing sector to navigate itself into a new position of `policy partners' in a broad public strategy, where housing, social and cultural integration and governance quality de ne the lead narrative. However, this does not mean that the resistance potential of Danish housing associations to the general trend of `dualization' of housing markets is in nite. In spring 2016, the Government launched a social security reform, which de facto impedes weak households from paying rents in some social housing estates, and an anxious debate about the lack of cheaper housing is now mounting. The strength of the social housing movement and the sustainability of the unitary model still depend on the mobilization of tenants, the capacity to perform the new role in a proactive and in uential manner and the market forces that attract more resourceful segments of the population to other tenure forms.
Whereas elected local tenants are the ultimate power base of the Danish system, the Swedish regime depends largely on its support from the SUT and its political legitimacy on the national level. Even after the signi cant transformation of the Swedish housing regime, it is still virtually impossible to nd a political actor in Sweden who stands up for market rents or for means tested social housing. Thus, the legitimate political ideal across the party spectrum in both countries still seems to be a coherent society in which dierent types of households and citizens are integrated. The question of the extent to which this political ideology can be sustained in practice given the on-going change in the unitary housing regimes remains.
Post scriptum. 'Ghetto problems' and 'particularly vulnerable areas'
In recent years the housing discourse in both countries has addressed social and be-havioural problems in a range of areas dominated by housing associations (DK) and municipal housing companies (SE). In Denmark, under the heading of 'ghetto problems', an annual list has since 2013 been issued by the government, singling out social housing areas with a combination of low income and education levels, poor labour market po-sitions, high criminal records and a high proportion of non-western immigrants. After previous 'ghetto plans' to deal with segregation in such areas, in May 2018 the Govern-ment launched a package of 22 proposals where the state directly regulates and monitors developments in a number of local estates (currently 30 areas). Targeting 2030, a
maxi-mum is set of 40 per cent social housing in designated areas, and the government retains a right to enforce sales and expropriation. On top, regulation covers tenant access to housing in these areas, depending on social status and criminal record. A double up of legal punishment for certain crimes in such areas is proposed alongside with demands for immigrant children to enrol in Danish day care. Furthermore, the nancial underpinning of the proposal initially threatened to drain NBF considerably. A broad majority in Par-liament stood behind the proposals, which mirrors a new political reality for the Danish social housing sector when the strong voter appeal of immigration, ethnicity and identity politics cuts across historical markers of political alliances.
Similar tendencies can be observed in the Swedish debate, but at least so far no such extreme political proposals have been launched, and the discourse is not only focused on MHC housing. However, the problems are framed in similar terms, and for some years now the Swedish Police have listed `particularly vulnerable areas' (currently 23) that have problems with poverty, criminality, low school performance, and high unemployment.
References
Abbas, P. (2002). Salg af almene boliger i Sverige. In P. Abbas, B. Mazanti, and Olsson, T. L., editors, Udenlandske erfaringer med salg af almene boliger. Geogra sk Institut, Copenhagen.
Annaniassen, T. (2013). Norge det socialdemokratiska ägarlandet. In Bengtsson, B., Annaniassen, T., Jensen, L., Ruonavaara, H., and Sveinsson, J. R., editors, Varför så olika? Nordisk bostadspolitik i jämförande historiskt ljus. Égalité, Malmö.
Arnott, R. (1987). Economic theory and housing. In Mills, E., editor, Handbook of regional and urban economics, volume 2, pages 959988. Elsevier, Amsterdam.
Bengtsson, B. (2001). Housing as a social right: Implications for welfare state theory. Scandinavian Political Studies, 24(4):255275.
Bengtsson, B. (2012). Path dependence or historical juncture? The Swedish housing regime before and after adjustment to the European competition policy. In ENHR 2012 Conference'Housing: Local Welfare and Local Markets in a Globalised World', Lillehammer, Norway, 2427 June 2012. Institutet för bostads-och urbanforskning, Up-psala universitet.
Bengtsson, B. (2013). Sverige-kommunal allmännytta och korporativa särintrssen (Swe-denpublic housing and corporatist special interests). In Annaniassen, E., Jensen, L., Ruonavaara, H., and Sveinsson, J. R., editors, Varför så olika? Nordisk bostadspolitik i jämförande historiskt ljus. Égalité, Malmö.
Bengtsson, B. (2016). Adjustment, resistance, or system shift?: The Swedish housing regime at the crossroads. In Uppsats till 5th Real Estate Workshop, Malmö 910 maj 2016.
Bengtsson, B., Annaniassen, E., Jensen, L., Ruonavaara, H., and Sveinsson, J. R. (2013). Varför så olika?: Nordisk bostadspolitik i jämförande historiskt ljus. Égalité, Malmö. Bengtsson, B. and Ruonavaara, H. (2010). Introduction to the special issue: Path
depen-dence in housing. Housing, Theory and Society, 27(3):193203.
Bengtsson, B. and Ruonavaara, H. (2011). Comparative process tracing in housing studies. International Journal of Housing Policy, 11(4):395414.
Bengtsson, B. and Ruonavaara, H. (2017). Comparative process tracing: making historical comparison structured and focused. Philosophy of the Social Sciences, 47(1):4466. Boligministeriet (1987). Den almennyttige boligsektors rolle på boligmarkedet.
Boligmin-isteriet, Copenhagen.
Boverket (2017). Allmännyttiga kommunala bostadsaktiebolag. Utvärdering av tillämpnin-gen av gällande lagstiftning. Boverket, Karlskrona.
CECODHAS (2011). State aid. http://www.housingeurope.eu/policy-actions/position-papers-eu-legal-framework/state-aid. (Accessed 01/07/13).
Crouch, C. and Farrell, H. (2004). Breaking the path of institutional development? Al-ternatives to the new determinism. Rationality and Society, 16(1):543.
DAB (2012). Et godt år for boligpolitikken. http://www.bl.dk/nyheder-og-presse/nyheder/et-godt-aar-for-boligpolitikken. (Accessed 02/07/13).
David, P. A. (1985). Clio and the economics of QWERTY. The American Economic Review, 75(2):332337.
Dynarski, M. (1986). Residential attachment and housing demand. Urban Studies, 23(1):1120.
Elsinga, M., Haner, M., and Van Der Heijden, H. (2008). Threats to the Dutch unitary rental market. European Journal of Housing Policy, 8(1):2137.
Elsinga, M. and Lind, H. (2013). The eect of EU-legislation on rental systems in Sweden and the Netherlands. Housing Studies, 28(7):960970.
Grander, M. (2015). Allmännyttans samhällsansvar. In Salonen, T., editor, Nyttan med allmännyttan. Liber, Stockholm.
Hall, P. A. and Taylor, R. C. (1996). Political science and the three new institutionalisms. Political studies, 44(5):936957.
Indenrigsministeriet (1945). Det fremtidige boligbyggeri. Betænkning afgivet af Inden-rigsministeriets Byggeudvalg af 1940. Indenrigsministeriet, Copenhagen.
Jensen, L. (2003). Den store koordinator: Finansministeriet som moderne styringsaktør. Djøf/Jurist-og Økonomforbundet.
Jensen, L. (2011). Introducing dialogue governance in Danish social housing: Opportu-nities and challenges to tenant democracy. In International Symposium on Social and Aordable Housing. 2011, Paper presented at a symposium at the Tsinghua University, Tsinghua, 710 June 2011.
Jensen, L. (2013a). Danmark-lokal boendedemokrati och nationell korporatism. In Bengtsson, B., Annaniassen, E., Jensen, L., Ruonavaara, H., and Sveinsson, J. R., editors, Varför så olika? Nordisk bostadspolitik i jämförande historiskt ljus. Égalité, Malmö.
Jensen, L. (2013b). Housing welfare policies in Scandinavia: A comparative perspective on a transition era. LHI Journal of Land, Housing, and Urban Aairs, 4(2):133144. Jørgensen, E. (1994). Boligforliget 1966. In Thomsen, B., editor, Temaer og brændpunkter
efter 1945.
Kemeny, J. (1995). From public housing to the social market: Rental policy strategies in comparative perspective. Routledge, London.
Mahoney, J. (2000). Path dependence in historical sociology. Theory and Society, 29(4):507548.
Nielsen, B. G. (2010a). The hidden politics of a haunted sector-retrenchment in Danish housing policy 2001-2009. Copenhagen University, Department of Political Science, PhD dissertation, Copenhagen.
Nielsen, B. G. (2010b). Is breaking up still hard to do? Policy retrenchment and housing policy change in a path dependent context. Housing, Theory and Society, 27(3):241 257.
Pagrotsky, S. (2010). Utmaningen av allmännyttan. Hur en svensk modell kan försvaras i en internationell värld. Stockholm University, Department of Political Science, Master thesis.
Pierson, P. (2004). Politics in time. History, institutions and social analysis. Princeton University Press, Princeton.
Sewell Jr, W. H. (2005). Logics of history: Social theory and social transformation. University of Chicago Press.
Sørvoll, J. and Bengtsson, B. (2018). The Pyrrhic victory of civil society housing? Co-operative housing in Sweden and Norway. International Journal of Housing Policy, 18(1):124142.
SOU (2008). EU, allmännyttan och hyrorna. 2008:38.
Stahl, K. (1985). Microeconomic analysis of housing markets: towards a conceptual framework. In Stahl, K., editor, Microeconomic models of housing markets, pages 126. Springer-Verlag, Berlin & Heidelberg.
Stephens, M. (2016). Using Esping-Andersen and Kemeny's welfare and housing regimes in comparative housing research. Critical Housing Analysis, 3(1):1929.
Stephens, M. (2017). Housing regimes twenty years after Kemeny. Working Paper. Heriot-Watt University, Edinburgh.
Streeck, W. and Thelen, K. (2005). Introduction: Institutional change in advanced politi-cal economies. In Streeck, W. and Thelen, K., editors, Beyond continuity. Institutional change in advanced political economies. Oxford University Press, Oxford.
Thelen, K. (1999). Historical institutionalism in comparative politics. Annual Review of Political Science, 2(1):369404.
Torgersen, U. (1987). Housing: the wobbly pillar of the welfare state. In Turner, B., Kemeny, J., and Lundqvist, L. J., editors, Between state and market: Housing in the post industrial era. Scandinavian Housing and Planning Research, Supplement 1.
