COMPARATIVE PLANNING AND HOUSING STUDIES BEYOND TAXONOMY : A GENEALOGY OF THE SPECIAL PROGRAMME FOR REHOUSING ( PORTUGAL )

Recent European comparative studies in the fields of housing policy and spatial planning have been dominated by taxonomical and linear approaches, and by normative calls for convergence toward systems considered more ‘mature’ or ‘advanced’. In this article, we adopt a genealogical perspective and consider those cultures that are central to the shaping of policy. We set out a long-term exploration of the intersection between spatial planning and housing policy in Portugal and focus on the Special Programme for Rehousing (Programa Especial de Realojamento, PER), a programme that has had changing roles (from a financial instrument to a core component of policies of urban regeneration) in connection with political and planning cultures changing in time and space. In this way, we provide evidence of the limited capacity of taxonomic and linear approaches to describe planning and housing systems undergoing processes of change and, conversely, show the potential of genealogical research.


Introduction
The Special Rehousing Programme (Programa Especial de Realojamento, hereafter PER) was launched in 1993 to provide financial instruments to the municipalities of the two main metropolitan areas of Portugal, Lisbon and Oporto, to rehouse the thousands of households living in informal settlements. 1 Despite having never been formally concluded, the PER had its acme during the late 1990s.Much had been written in that decade, but the PER and housing issues seem to have fallen out of fashion since then.Indeed, in late 2014, while preparing the bid for the research project for the project from which this article stems, we found among fellow Portuguese scholars a lack of interest in the PER, and housing policy more generally.In the wake of the global economic crisis, issues of housing once more took centre stage (see Garcia-Lamarca and Kaika, 2016;Madden and Marcuse, 2016).Paradoxically, it was the recovery, rather than crisis and subsequent austerity politics, that triggered the debate about housing in Portugal, and particularly the gentrification and 'touristification', which followed the boom in tourism and real estate sectors.In 2016, the UN Special Rapporteur for Adequate Housing visited Portugal and highlighted the permanence of precarious housing conditions in informal settlements (Fahra, 2017).In March 2017, the national parliament urged the government to evaluate housing needs and act.
Recent academic and political attention to housing has reverted a long trend, started around the 1970s/1980s, of declining public intervention in housing policy and the commodification of housing (see Madden and Marcuse, 2016).The commodification of housing, however, happened at the same time as the focus of public policy shifted toward promoting urban regeneration, hence making housing one component, among many others, of planning policy.This article intends to explore these transitions by focusing on the intersection between the European fields of comparative planning and housing studies.Both fields have been dominated by taxonomic and linear approaches that have often proved unable to provide an in-depth understanding of the dynamics of policymaking in areas that are characterised by complex intersections between different sectorial policies and levels of governmental action.By adopting a genealogical approach (see Foucault, 1994Foucault, [1976]]; Gutting, 1990) and using the concepts of 'planning cultures' (Knieling and Othengrafen, 2015) and 'public action' (Allen et al., 2004), we intend to gain a better understanding of the development of the policy process -in a way that unites the characteristics of the local environment, the dynamic of overarching trends and the process of policy diffusion, and the complex network of relations between actors sitting at different scales.This theoretical approach is, we argue, especially needed in the case of Portugal.This is for two reasons.First, comparative exercises have so far failed to define the country's planning system in the light of conventional taxonomic categories (see Campos and Ferrão, 2015).Secondly, reflecting on Portugal's awkward place in comparative planning research (and, more broadly, on the country's position at the 'borderlands' of urban theory, see Baptista, 2013;Tulumello, 2016b) can enrich and de-parochialise our theoretical understanding of the phenomena under observation -such as the idea of 'sociology at world scale' proposed by Robinson (2011).
Our theoretical goals are to provide evidence of the limited capacity of taxonomic and linear approaches to describe planning and housing systems, and especially their evolution over time; and to show the greater potentiality of a genealogical approach.We will pursue these goals by locating the specificities of the PER in the Portuguese socio-political context -with particular attention to the political cycles of national governmentand, at the same time, by reflecting on the impact of the programme on the Portuguese planning and housing system.
There are two main reasons for us to study the PER.First, it is a good example of a multilevel policy without territorial governance -the national government provided the funding instruments, financial regulations and generic guidelines, while municipalities were in charge of planning, design and implementation.Secondly, the case of the PER allows us to explore the tensions that exist between the dimensions of housing provision and urban regeneration.The PER constituted an incipient regional plan: it significantly restructured the regional built environment -in the Lisbon metro more than 20,000 households were rehoused -but it did so in the Transactions of the Association of European Schools of Planning • 2 (2018) 32-46 absence of proper planning instruments.As such, the extent to which the provision of housing was integrated with larger concerns pertaining to urban regeneration was dependent on contingencies, contextual factors, and local trajectories of path dependency, including, for example, the approaches to planning deployed by each municipality, and the different phases of national governmental action.
The article starts by emphasising the dominance of taxonomic and linear approaches in comparative planning and housing studies (Section 2).We then set out our genealogical perspective for the study of the Portuguese case (Section 3).The backdrop of the evolution of the Portuguese planning and housing policy is explored in Section 4, before we present the history of the PER with specific reference to its implementation in Lisbon metro (Section 5).We conclude by summarising the contribution that our genealogical exploration of the Portuguese case brings to comparative studies more broadly.

A Brief Summary of the European Field of Comparative Planning and Housing Studies
European comparative studies have tended to focus on two key elements: first, within a taxonomic perspective, they have investigated the differences that exist in the various national planning and housing systems; secondly, they have investigated the process of convergence of these systems, for example in the light of the so-called 'Europeanisation' (Giannakorou, 2005;Ferrão, 2011).
The Compendium of Spatial Planning Systems (CEC, 1997), commissioned by the European Commission, triggered a rush of comparative planning studies (see, amongst others, Farinós Dasí, 2007;Nadin and Stead, 2008;Stead and Cotella, 2011).The Compendium divided planning systems of EU member states in four 'traditions' (CEC, 1997, pp.33-37): the regional economic approach (typical of France); the comprehensive integrated approach (Denmark and Netherlands); land-use management (UK); and urbanism (Italy and Spain); with other systems falling in between two or more traditions.The Compendium, and most following comparative works, have discussed the 'maturity' of systems to foster (linear) convergence toward those considered more mature.Maturity was broadly defined in terms of the 'degree of public acceptance of the need for planning', of the 'provision of up-to-date policy instruments', of the 'degree of vertical integration and cooperation between levels of administration', and of the 'existence of transparent and productive consultation mechanisms' (CEC, 1997, p.35).That a major ongoing ESPON comparative study is concerned with measuring the maturity of national systems (Nadin et al., 2016, p.10) suggests that the dominant approach remained quite stable over time.
Comparative housing studies (see, for different perspectives, Harloe, 1995;Kemeny, 2001;Alves, 2017;Di Feliciantonio and Aalbers, 2017) have focused on two main dimensions: the composition of housing systems (e.g.types of tenure, weight of public/private/non-profit sectors, quality of the built environment); and (national) policy approaches.Within this field, many have been concerned with emphasising the systems 'lagging behind', typically Southern European ones, in the implementation of housing policies when compared with more 'advanced', typically central European, countries (Allen et al., 2004).At the same time, most studies have described a process of convergence within housing policies since the 1980s based on the neo-liberal transformation (and retrenchment) of welfare states, that is, the shift from state-provided social-rented housing towards regulation, the stimulus of home ownership, the support given to private/charity social housing, and the privatisation of public housing stocks.Doling (2006) suggested that the EU has played a role in this transition, developing a deficient housing policy 'by stealth', made up of regulations in different fields.All in all, the increase in levels of home ownership and private renting (and therefore the shrinking of the social housing component) seems to be 'desirable' for EU policies/politics (Doling, 2006) 2 .In the context of state withdrawal from direct housing provision, public resources have been re-directed toward the regeneration of the built environment (Cameron, 1992), as opposed to the emphasis on new developments that was prevalent during the post-WWII golden age of public housing provision.

Beyond Taxonomy and Linearity in Comparative Studies: Epistemological and Methodological Notes
The fields we have surveyed in the previous section have, on the one hand, provided a set of 'boxes' (e.g.'traditions', 'paradigms'), where systems can be placed and compared.On the other, they have captured grand processes of transformation at the European level; namely, the Europeanisation of spatial planning and the neo-liberalisation of housing policies.Recently, critiques of the dominance of taxonomic and linear approaches have emerged in the fields of planning and housing studies.In planning theory, doubts have been cast over the explanatory capacity of the concepts of paradigm and tradition (Getimis, 2012;Tulumello, 2015).The field of planning cultures emerged to contribute to the de-parochialisation of planning studies (Sanyal, 2005) in line with post-colonial approaches to urban studies (Roy, 2009;Robinson, 2011).In Europe, Knieling and Othengrafen (2015) developed the 'culturised planning model' to explore the trajectories of planning policy by looking into 'taken for granted assumptions' and 'unwritten patterns of power ' (p.2135).With respect to the harmonisation of planning systems amid Europeanisation, Knieling and Othengrafen concluded that adaptational pressures, for example Europeanization, obviously result in the customization of existing structures, frames and policies ('planning artefacts' and 'planning environment') but do not necessarily touch the underlying core cultural traits ('societal environment') (2015, p.2144).
That is why, they argue, static comparisons focusing on institutional and formal features of the planning systems often fail in explaining specific spatial developments across EU member states (Knieling and Othengrafen, 2015; see also Fürst, 2009, p.27).This approach resonates with the concept of 'public action' (Allen et al., 2004) in housing studies, namely the idea that we should look at housing issues and housing policy 'as the outcome of a system of relationships among the different actors involved ' (p.57).What the approach of planning cultures and public action have in common is the critique of the tendency of taxonomic and linear approaches to project normative ideals in contexts where these may not be adequate or even relevant.This has two implications: from an analytical perspective, there is a risk of losing sight of the peculiarities of specific cases, particularly with regards to processes of policy change; from a normative perspective, there is a risk of reproducing and imposing hegemonic ideals outside the contexts where they have developed and are fully meaningful.
In contrast, the epistemological approach of this article is the adoption of a genealogic perspective, which is applied through an account of the history of the PER against the background of the evolution of the Portuguese planning system and housing policy.As a methodology of historical research, genealogy (Foucault, 1994(Foucault, [1976]]; Gutting, 1990) is designed to investigate the non-linear nature of the dynamics of societal change -from one system of thought and knowledge (a 'discursive formation', in Foucault's words) to another.Foucault set out three key methodological elements for genealogical investigations: specificity, discontinuity and exteriority.The idea of specificity constitutes an exhortation to describe each discursive formation on its own terms, acknowledging its contingent and self-sustained nature, and its inherent logic -in contrast to an ahistorical analysis based on supposedly universal categories, ex-post judgments, and so forth.The idea of discontinuity suggests that we should investigate ruptures and mutations (or the absence thereof) that intervene between successive discursive formations (and policy formulations).Finally, the concept of 'exteriority' calls the researcher's attention to the material, practical conditions of existence of each discursive formations -suspending, so to speak, the tension towards the search for deeper meanings and interpretation.
In this article, the genealogic approach is deployed in order to understand the dynamics of change in the particular system of power and knowledge represented by the field of Portuguese planning system and housing policy in the last 30 years -the PER being our entry point.The contribution that the adoption of a genealogical approach to a single case can provide to the field of comparative research can be summed up in two steps.First, one single case can be sufficient to call into question conventional wisdom (Flyvbjerg, 2006), for instance, by showing how a taxonomy cannot accurately describe that very case.This is particularly relevant when studying contexts that have long remained at the 'borderlands' of mainstream theorisation (Baptista, 2013;Tulumello, 2016b).Secondly, the analysis of a single case study over a long time-span can produce more nuanced theorisations and avoid recourse to shallow culturalist explanations of the divergence of certain national contexts from normative ideals produced elsewhere (see Tulumello, 2016a, pp.6-7).This allows us, in turn, to avoid falling into the trap of particularism, and look instead at processes of continuity and change in a given policy area by accounting for the mutual influence between local context and wider structural processes.
Our reconstruction of the history of the PER is based on three sources of evidence: first, the original policy documents (laws, decree-laws, diplomas, reports, governmental programmes); secondly, in-depth interviews with key informants3 ; and thirdly, field notes from several meetings with civil servants in most municipalities of Lisbon metro during the collection of data for the research project4 .

Before the PER: The Portuguese Planning System and Housing Policy between Continuity and Change
Placing the Portuguese planning system and housing policy within existing taxonomies is no easy task.Campos and Ferrão (2015) note that there is no consensus in comparative literature with regard to the understanding of the Portuguese system -the key features of which are linked by the Compendium, for example, to three out of four planning 'traditions' (CEC, 1997).Portugal's awkward position seems to reflect the broader problem of fruitfully including 'Southern perspectives' (Janin Rivolin and Faludi, 2005) in the EU debate on planning amid the dominance of the maturity perspective.This section attempts to describe the main traits of the Portuguese planning system and housing policy before the launch of the PER in 1993.Campos and Ferrão (2015) argue that planning has traditionally had a weak status in Portugal.Specifically, they highlight three key historical features: i) the dominance of architects and engineers in the field (hence the link to the tradition of 'urbanism', as per the Compendium's definition; CEC, 1997); ii) the lack of a sound reference framework for spatial planning at the regional level; and iii) the centralised nature of policymaking in this area, combined with the state's limited effectiveness in confronting the interests of the real estate sector.The authors also describe the long and significant process of change that has occurred in the last three decades.In the aftermath of the 1974 democratic revolution, the creation of municipal power inaugurated a dialectic with the centralism of the national government.The decade that preceded the enactment of the PER (1993) and the years that immediately followed it were marked by significant steps toward a more integrated planning perspective at the urban and regional level.In 1982 the municipal masterplan (Plano Director Municipal, PDM) was introduced in planning regulations; virtually all municipalities had produced their own local master plan by the early 1990s when municipal masterplans became a mandatory requirement for municipalities to apply for EU funds and expropriate land (Campos and Ferrão, 2015, p.18).The municipal masterplans 'introduced into public consciousness the idea that urbanisation and construction had to follow rules […] [and] created, for the first time, a market for specialists in urbanism and spatial planning' (Campos and Ferrão, 2015, p.19).
Different factors contributed to this trend: i) a period of relative political stability after turbulence which had marked the first years after the revolution; ii) the opening, in 1982, of the first Portuguese master's programme in urban and regional planning in Lisbon; iii) the development of a new interdisciplinary approach, which conceived housing as a key component of urban planning (former Secretary of State for Spatial Planning and Cities, interview); iv) the role of geographers, during the 1990s, in 'the establishment of spatial planning perspectives as a counterweight' to the traditional urbanism approach (Campos and Ferrão, 2015, p.19); and v) Portugal's adhesion to the European Community in 1986, with participation in the European Spatial Development Perspective (ESDP).New planning instruments and bodies, inspired by a more integrated and strategic approach to spatial planning, were introduced with the goal of consolidating the planning system.A comprehensive reform (Spatial and Urban Planning Policy Act, Law 48/1998) regulated spatial planning as an autonomous policy area; introduced a clear-cut distinction between the statutory and developmental arms of the planning system; and harmonised Portuguese planning law with key principles derived from the ESDP process.The Portuguese housing system is similar to other Southern European ones (Allen et al., 2004), especially with regard to the high levels of private ownership and low levels of social rented housing; the high proportion of second homes; the crucial role of family welfare in supporting access to homes; and the significant role of selfpromotion.This configuration has emerged over time due to three main factors, namely i) a political consensus on home ownership as an effective way to achieve social sustainability; ii) the sale of public rented housing stock as a result of institutional difficulties to manage it; and iii) an orientation of public administration toward 'implementing rules to control private initiative, rather than developing "entrepreneurial attitudes" within the public sector' (Allen et al., 2004, p.166).Furthermore, housing has been traditionally considered (especially at the level of central administration) as an issue of financing and public works -'giving people a home' (dar casa às pessoas) in the words of many of our respondents5 ; and acted upon through single programmes rather than a comprehensive strategy.In the words of a Professor in Sociology and Urban Studies (interview), there has never been a housing policy in Portugal.There have been packages [pacotes] […].Why packages?Because one can control goals, and funds allocated to it, better.Now, an integrated housing policy, thinking of housing needs and solutions, this is something… Other respondents shared this idea.A former Secretary of State for Spatial Planning and Cities, for example, pointed to the many institutional and political pushbacks encountered during one of the most ambitious attempts of developing an integrated housing policy (see Section 5.4).These included i) the difficulty in transforming the main governmental agency in the housing sector (the Institute of Housing and Urban Regeneration, Instituto de Habitação e Rehabilitação Urbana, IHRU) from a financial institution to a promoter of housing/regeneration policy6 ; ii) the bipartisan consensus on the importance of home ownership; iii) the processes of alienation of public housing stock already ongoing at the time; iv) and the failure in passing a national strategy for housing policy.Another respondent (researcher and former IHRU board member), pointed out that the Portuguese experience in recent decades has followed a path, in which some ideas, generated more or less consciously during earlier eras laid the conditions for further advancement.
All in all, however, the historic trajectory of Portuguese housing policy significantly contributed to the buildup of the housing crisis of the early 1990s, which the PER was enacted to solve.During the dictatorship era , the primacy of private intervention coupled by scarce and paternalistic public interventions (Serra, 1997) made the housing system unable to accommodate population growth and rural-urban migrations.After the 1974 democratic revolution, housing issues were brought to the limelight by a wave of activism and the 'right to housing' was included in the democratic Constitution (art.65).With the creation of local power and processes of decentralisation, housing came to be considered a de facto responsibility of municipal authorities, a proximity policy to be developed in close relation with urban planning.In the late 1970s several programmes to promote social housing and improve informal settlements were launched (see Bandeirinha, 2007).Although these represented the first attempts at making housing policy a core element of urban policy (researcher and former IHRU board member, interview), they did not bring about structural transformations.Indeed, housing problems (including the proliferation of informal settlements in the metropolitan areas) worsened after the revolution, as housing policies and the market were unable to accommodate the housing needs of hundreds of thousands of immigrants from the former African colonies -both Portuguese settlers with their descendants and people of African descent.

The Programa Especial de Realojamento: Genealogy of a Housing Policy
This section describes how the PER came about; reflects on the implications of the program on the housing situation and urban growth; and discusses the extent to which the programme has been a component of wider attempts at urban regeneration or remained a narrow, sectorial housing policy.
5.1.Before 1993: Early Policy Responses to the Housing Crisis and the Birth of the PER Before 1993, some policies for the regeneration of the built environment were launched, but they still fell short of solving the problem of informal settlements.In 1985 the government launched the Programme for Urban Refurbishment (Programa de Reabilitação Urbana, PRU).This expanded to public spaces the interventions on degraded housing that had taken place under the Programme for the Recovery of Degraded Buildings (Programa de Recuperação de Imóveis Degradados, PRID).The PRU increased financial resources and was an important attempt at integrating housing policy in the broader framework of urban policy.The PRU, not being restricted to intervention in specific areas, could theoretically cover interventions in any degraded area, including illegal neighbourhoods.However, a few months later, Decree-Law 366/1985 (later replaced by Decree-Law 226/1987) provided special guidelines for rehousing populations living in informal settlements, de facto separating interventions in these areas from regeneration.
With the beginning, in 1985, of a decade of centre-right governments (which largely conceived housing as a 'productive sector', and the purview of private companies) regeneration became marginalised to planning policy and rehousing was considered the remit of local authorities, with the state providing some financial support (Decree-Law 226/1987).Towards the end of the 1980s, the housing shortage was still a major problem, for a plethora of reasons (Ferreira, 1988, pp.55-56).These included i) the absence of a comprehensive national housing policy7 ; ii) the relevance of land property in processes of accumulation -a typical feature of Southern European countries (Salzano, 1998); iii) weaknesses within the financing system; and iv) scarce technological and organisational development in the building industry.The electoral campaign of 1991 seemed to mark a turning point: pressured by public opinion and opposition parties, the prime minister promised more incisive action in housing.This promise, however, remained forgotten until 1993, and public investment in housing policy remained stagnant until that date.Overall, public action was not capable of taking the lead in regulating urban growth and the promotion of housing (Serra, 1997).
Within the contradictory trends that we have described, pressure mounted, in the early 1990s, for decisive political action to solve what was perceived to be a pressing social and urban concern.The urgency was the product of several interlocking factors.The problem of informal settlements in Lisbon had reached unprecedented levels.Even more importantly, a more acute awareness began to develop among activists, decision-makers, and public opinion at large.Portugal's membership of the European Community and the organisation of a number of high-profile international events (and particularly the 1998 World Fair) added to the urgency.The state of Lisbon's peripheries gained media attention (see Alves, 2016); several activists' campaigns against evictions received considerable exposure; and the informal settlements became an object of political campaigning, with the mayor of Lisbon and the President of the Republic -both affiliated to the Socialist Party, the main opposition to the centre-right government -promoting high-profile public initiatives designed to put pressure on the executive to address the problems.'Acabar com as barracas' ('getting rid of the shacks') became the largely consensual formula that propelled the enactment of the PER.
A few months before the enactment of the PER, an expert committee, made up of scholars and practitioners politically close to the Socialist Party, was formed through the initiative of various organisations operating in the field of housing.The committee published the influential White Paper on Housing Policy in Portugal (Livro Branco sobre a Política de Habitação em Portugal; Ferreira, 1993).It called for a plan for the construction of 500,000 houses (p.67), and the creation of a publicly owned stock of land to be allocated to the plan (p.70).Some 60,000 of those houses were indicated as instrumental to the 'provision of socially rented housing on an appropriate scale, in order to allow for the elimination of the informal settlements and illegal allotments existing in the metros of Lisbon and Oporto' (p.60).This stock was supposed to adopt an investment and management model (pp.60-68) characterised by: i) partnerships between public authorities (local, national or regional housing companies to be funded by the central administration and the municipalities) and residents' associations; ii) the adjustment of rent levels to market value; iii) the refurbishment of public housing stock; and iv) the possibility of selling public housing units to tenants.The White Paper further suggested the integration of programmes for the refurbishment of existing public housing stock into a wider refurbishment programme including 240,000 housing units (p.22), which was also expected to contribute to reducing the need for social housing developments.In short, the White Paper sought a systematic integration of the goals of housing provision and urban regeneration, an integration neglected by subsequent policies launched by the centreright government.

1993-1995: PER as a Purely Financial Instrument
In 1993, amid growing public and political debate about housing issues, the government launched six measures, announced as a comprehensive 'housing package' (pacote da habitação).The Decree-Law 163/1993, which established the PER, indeed intended to 'get rid of shacks'.It followed the suggestions presented in the White Paper in relation to the cooperation between central and local authorities, and the adoption of a system of subsidised rents.However, differences between the Decree-Law and the White Paper were significant and, notably: i) in relation to the nature of the financial resources, the White Paper's suggestion for the creation of a specific national fund made up of both national and EU resources was discarded; ii) while, in addition to new developments, the White Paper focused on refurbishment, the PER envisaged the possibility to purchase on the market; and iii) the Decree-Law did not provide any guidelines as to the management model.The PER was not very dissimilar to previous legislation, especially Decree-Law 226/1987.Its novelty consisted primarily of the better financial conditions offered to the municipalities with regard to borrowing funds, and the overall availability of more finance8 .The PER intended to solve the 'plague' (Decree-Law 163/1993) of informal settlements in the two main metropolitan areas of the country.According to the then Minister for Public Works, Transport and Communication, who was responsible for the launch of the PER (interview), the government delegated implementation entirely to the municipalities in order to privilege effectiveness: had any sort of coordinating structure been put in place, this would had signalled an intention of the state to 'impose its will' over the municipalities, creating a conflict that would complicate the implementation of the programme.
The emphasis on the complementarity between new construction and the purchase of housing units on the market is an illustration of the centre-right government's housing policy.The PER aimed at stimulating the 'contribution of the market' for the development of the social housing sector.For this reason, especially important was Decree-Law 165/1993, which reformed the legal regime of the contracts signed between public institutions and private companies for the construction of housing units, making it easier for private developers to participate in the development of social housing sector.
The land policy, one of the cornerstones of the White Paper (Ferreira, 1993, pp.74-76), was completely forgotten in the 'housing package'.Although another Decree-Law (164/1993) acknowledged the 'lack or relative scarcity of land that would allow construction at affordable prices', interventions on this issue remained limited to allocating publicly-owned land for construction -and included no attempt to counter speculation in the real estate sector.Similarly, the only aspect of 'regeneration' included in the PER was the obligation to demolish the shacks once the rehousing had been completed.
In the words of a respondent, the former Secretary of State for Spatial Planning and Cities, the PER was a financing instrument based on an agreement with local authorities.What happened is that some local authorities had a more strategic vision and included the PER in wider strategy of [urban] interventions, others [did not] […] Let us be crystal clear, the PER was a funding instrument.Nothing more than that.The ways it was adopted by each local authority may have been different […] Virtually no one in [the field of] spatial planning considered the PER a spatial planning instrument.
The design of the PER as a financial solution to an exceptional problem was consistent with the long-term shift of Portuguese housing policy away from housing provision; the transition was wholly and fully completed during the decade of centre-right governments (1985)(1986)(1987)(1988)(1989)(1990)(1991)(1992)(1993)(1994)(1995).The building sector was liberalised, and this paved the way for the boom in real estate investments and construction after Portugal joined the European Community.Since then, public housing policy was de facto promoted by the Department of Finance, and consisted primarily of incentives to home ownership (Santos et al., 2014) -a 'concealed' housing policy, according to a former Secretary of State for Housing (interview). Overall, between 1987 and2011, 17.9% of public expenditure in housing was used for provision (basically, the PER), while 73.3% was funnelled to subsidies for home ownership (IHRU, 2015, p.4; see Figure 1).The housing sector and housing policy played a crucial role in the financialisation of Portuguese society and its 'semi-peripheral' economy (Santos, 1985;Santos et al., 2014). 5. 3. 1995-2002: Steps toward an Urban Policy?

Rehousing programmes
During the first few years, due to the financial nature of the PER, and the total autonomy of municipalities, its implementation was virtually exclusively dependent on local planning cultures and political traditions.
Despite the great expectations that the launch of the PER generated among experts and civil servants (Cachado, 2013), the PER did not gain pace during the first few years (see Figure 1) except within municipalities such as Lisbon and Oeiras, which had already developed their own rehousing programmes.In addition, the capacity of the PER to be integrated within wider urban policies was quite uneven.In 1997, only a third of the municipalities within the Lisbon metro had included urban refurbishment and regeneration among their goals and 'only a fraction of the municipal authorities involved have articulated the PER within broader strategies of territorial regeneration' (Guerra, 1999, p.60).
The new socialist governments (1995-1999; 1999-2002) prioritised speeding up the implementation of the PER, while at the same time they also tried to make it a more flexible and comprehensive policy instrument.For the first time in the history of democratic Portugal the housing portfolio was placed in a ministry responsible for urban and regional planning, that of Infrastructure, Planning and Territorial Administration.This innovation was short-lived as the next executive reformulated the organisational chart and re-separated the portfolios of planning and housing.The Secretary of State for Housing at the time (interview) maintained that virtually no rehousing operations were in fact conducted under the previous government.New regulations were, therefore, approved to make the PER more flexible9 , and the rate of expenditure accelerated swiftly, peaking in 2000 (Figure 1).In 1996, the PER-Households (PER-Famílias) was introduced to allow individual households to use the financial instruments of PER to enter the private housing market.
In the absence of local authorities acting beyond simply providing housing, the government used EU Structural Funds to complement the rehousing process with the provision of public services and businesses (former Secretary of State for Housing; civil servant in the national Department of Housing, interviews).In 1996, the sub-programme Operational Intervention for Urban Renewal (Intervenção Operacional Renovação Urbana, IORU) was reformulated to allow for the construction of open spaces and public facilities, and to support the establishment of small private enterprises (see also Coutinho, 1997).By widening the offer of programmes that could complement rehousing10 , the socialist government established a more integrated approach: The socio-economic integration of the rehoused families was now seen in the context of the urban development.
Regeneration became, however, dependent on EU funds, which funded most complementary programmes implemented by municipalities (Guerra, 1999).
Guerra observed that the PER had no significant influence on planning policies in the Lisbon metro, for two reasons (1999, p.60): first, improving informal settlements was not a central goal of municipalities; and secondly, the PER was implemented when the municipal masterplans had been already defined.Indeed, in half of the municipalities, the PER only arrived during the late phases of the elaboration of the masterplans.However, in the other half of the municipalities, the two processes developed together (masterplans approved between 1995 and 1999).Following Guerra (1999, pp.71-72), in 1997 the municipalities that were developing rehousing operations of an average scale presented the highest rate of progress (>25%).This observation led Guerra to conclude that the presence of either a relatively large or a relatively small number of households to be rehoused was a factor in delaying the action of municipalities (1999, p.73).We wonder, however, whether the pace of implementation was not accelerated, at least in some cases, by the lack of a masterplan in forcea case in point could be that of Sintra, a municipality that implemented medium-scale rehousing operations and attained, in 1997, the highest rate of progress in the northern part of Lisbon metro (46.20%, against a national average of 20%).The impact of local planning regulations also deserves more attention in light of the publication, in 1997, of Decree-Law 156/1997, which increased the flexibility of municipal land use regulations in areas interested in rehousing programmes, creating a sort of state of exception for implementing the PER because it allowed municipalities to swiftly change land uses, building regulations and development plans.While it is difficult to estimate the impact of Decree 156/1997 on the implementation of the PER, the years following the decree were indeed those with the most budgeted operations (Figure 1)11 .

2002-…: Demise and Rebirth of the PER
A new phase of the PER story began in 2002 with a change of government from centre-left to centre-right.The program of the new government adopted a discourse that was in line with the prevailing emphasis on integrated and sustainable urban development in European debates.The key objective of housing policy became the creation of incentives for the refurbishment of existing buildings and the revitalisation of the rental market.As far as social housing was concerned, the main goals became the definition of policy instruments for the renewal of the existing housing stock and ensuring the progress of the PER.
The discourse of the new 'urban policy' (política para as cidades), in line with wider and academic debates (Guerra et al., 2005), had urban regeneration at its core -and for the first time a ministry for urban policy was created (Ministry of Cities, Spatial Planning and Environment).However, in line with the governmental goal to 'reduce the role of the State in the development and management of housing provision' (Decree-Law 199/2002), virtually no national resources were invested in this field (Figure 1; IHRU, 2015).Housing policy, with the exception of the programme POLIS, remained, therefore, almost completely dependent on EU funds and private investments (civil servant at Lisbon Municipality, interview).The PER followed this pattern.Shifting away from construction and toward regeneration, a new set of amendments (e.g.Decree-Law 271/2003) eased the selling of municipal housing units built with the PER or other municipal programmes.In addition, and for the first time, municipalities were allowed to use PER funds to build public facilities and services.In 2004, the PROHABITA programme was launched (Decree-Law 135/2004).This provided a broader framework for rehousing interventions outside of the metros of Lisbon and Oporto and extended the provision of the law to all situations where there were severe housing stresses.
Despite the introduction of significant conceptual, legal and operational changes, the national budget approved for 2003 introduced a set of limitations on municipalities' ability to borrow funds.While it did not affect the projects already budgeted, the new provision made it virtually impossible to contract new rehousing operations.Concurrently, stricter budgetary requirements pushed the municipalities to sell council housing stock to fund new developments.Even before the passing of the 2003 state budget, various municipalities had voiced their concerns about the possible demise of the PER (see Morais, 2002).This proved to be the case after 2003; expenditure for rehousing plummeted (Figure 1) -and there was a parallel rise in critical voices (see Felner, 2003;Simão, 2003).
The Socialist Party took back control of the government in 2005, and promoted the most comprehensive attempt to structure housing policy within urban and regeneration policies.Responsibility for housing was given to the Secretary of State for Spatial Planning and Cities, a renowned Professor of Geography.The attitude of the government was signalled by the transformation of the National Institute for Housing (Instituto Nacional da Habitação, INH), basically a financial institution, into the Institute for Housing and Urban Regeneration (Instituto da Habitação e Reabilitação Urbana, IHRU).At interview, the former Secretary of State stressed that his goal was transitioning from a 'social housing policy' (politica de habitação social) toward a 'social policy for housing' (política social de habitação) -i.e.articulating housing provision within urban regeneration.
With the goal of integrating housing policy into the broader framework of urban and regional planning, the government vowed to conclude the implementation of rehousing programmes, and to promote a policy of regeneration within and beyond these programmes.Two initiatives were launched.First, an amendment to PROHABITA (Decree-Law 54/2007) was introduced to address households that were living in informal settlements that were scheduled for demolition in the context of the PER, but had not been originally included in the survey of 1993.These households were, and in some contexts still are (as denounced by activist campaigns), facing evictions for the advancement of slum-clearances.Secondly, a number of policies and programmes for urban regeneration -including the Critical Neighbourhoods Initiative (Iniciativa Bairros Críticos, IBC) and the Urban Regeneration Partnerships (Parcerias para a Regeneração Urbana, PRU) -were launched.However, these were never fully implemented as a consequence of the global financial and economic crisis of 2010.All in all, as the former Secretary of State for Spatial Planning and Cities, (interview) admitted, the failure to pass the national strategy for housing policy negated most reform attempts.In 2011, the socialist government was forced to ask for international financial help, and on June 2011 a centreright government took office.The new government -amid the pressures of the external borrowers -enforced a harsh austerity agenda (Pedroso, 2014).This placed housing at the margins of governmental agendas.As of today, according to one of our respondents, a Professor of Urban Sociology, housing policy is exclusively 'local', in that no state intervention exists.Against the background of various, if often contradictory, attempts at decentralisation, urban policy has become an important focus for local government (see Campos and Ferrão, 2015).However, rather than political competences and resources, the central government has been delegating 'duties and obligations' (Mendes, 2016), making urban policy much dependent on an uneven distribution of resources and skills among municipalities.
In the absence of funds, rehousing programmes progressively ceased while demolitions and evictions multiplied (Alves, 2016).This was especially true in the municipalities of Amadora, Loures and Seixal.At the same time, after the negotiation of the memorandum of understanding with the external borrowers, different reforms were passed to liberalise both the rental market and the planning system.Only after the return of the Socialist Party to the government in 2015 (as majority stakeholder of a parliamentary coalition including the Left Bloc and the Portuguese Communist Party) has a new phase for Portuguese housing policies appeared to be opening.Amid political and public pressures following the extremely critical report on local housing conditions by the UN Rapporteur for Adequate Housing (Fahra, 2017), the government appointed a new Secretary of State for Housing, and has launched the New Generation of Housing Policies (Nova Geração de Políticas de Habitação, NGPH).The latter includes the programme First Right (Primeiro Direito) that provides financial instruments to municipalities for addressing 'serious housing needs' (Decree-Law 37/2018).At the same time, a draft for a framework law for housing (Lei de Base de Habitação) is being discussed in the Portuguese parliament.Twenty-four years after its launch, and a decade or so since its de facto demise, the PER is reborn from its own ashes, and is itself triggering renewed political attention and policy action with regard to housing policy in Portugal.

Conclusion
What emerges from the history of the PER, seen against the backdrop of the planning/housing system(s), is 'something one might call a genealogy, or rather a multiplicity of genealogical researches, a painstaking rediscovery of struggles together with the rude memory of their conflicts' (Foucault 1994(Foucault [1976]], p.83).The historic reconstruction has shown how the Portuguese planning and housing fields are characterised by complex patterns of continuity and change, marked by inertia and sudden accelerations, by critical junctures and watershed moments, as well as by tensions between different levels of government and policy traditions.
We have been especially interested in exploring the intersection between a 'weak' planning system in transition, and the convergence of housing policy toward home ownership and regeneration.These relations seem to be particularly problematic for a number of reasons (Carmo et al., 2014): the limits of statutory instruments which are incapable of regulating urban growth but at the same time too rigid to allow for the regularisation and upgrade of informal settlements and illegal allotments; and the dependency of municipal authorities on the issuing of building permits as a source of financing -within a context of decentralisation without resource transferences.In this respect, the case of the PER allows us to add some nuances to consolidated ideas about housing and spatial planning in Portugal, as well as policy diffusion in Europe.
In the first place, the PER largely failed to establish itself as an example of genuine urban policy.Despite the promise noted by the early 1990s -with the development of a more thorough reflection on housing policies and the introduction of significant novelties in the planning system -the design and implementation of the PER were consistent with long-standing characteristics of previous housing and planning policies.Three main factors prevented the institutionalisation of the programme into a plan with strategic vision over urban and regional development, and cogent regulatory power.First, the programme lacked coordination at the regional level.Secondly, this was because of the institutional nature of the actors involved at the central level, seen, for example, in the lack of interest at the IHRU, an essentially financial institution, for territorial dimensions, and the local level, for instance in a lack of planning professionals and technical expertise in several municipalities.Finally, the programme was strongly dependent on electoral cycles.While this seems to support the thesis of an enduring 'immaturity' in the system, there are some other elements that contradict this idea.First of all, the design and implementation of the PER was influenced by a number of contingent factors, particularly a diffused sense of urgency around the issue of the informal settlements.The 'quick-and-dirty' nature of the programme (as opposed to a more holistic and integrated approach) in a way represented an effective answer to the problem.Secondly, the PER itself evolved over time, gradually moving toward a more integrated and flexible design.Thirdly, despite the awkward, vintage nature of the PER -a large programme of direct housing provision implemented when most European states had already abandoned this kind of intervention -Portuguese housing policies proved relatively up-to-date in following broader continental trends.Indeed, since the 1980s the largest share of Portuguese spending for housing consisted of subsidies for home ownership, boosting the financialisation of housing.Finally, the political dynamics and irregular flow of financial resources further blur the contours of the PER in terms of its impact on local cultures in the fields of planning and housing policies.A notable example is that of the reforms by the centre-right government in 2002/2003, which did indeed make the PER more flexible, by allowing its use to provide public services; at the same time, however, the government virtually killed the programme by halting the flow of financial resources available to it.
To conclude, we suggest that traditional and taxonomical explanations of the relative 'underdevelopment' of Portuguese (and Southern European) planning and housing policies -which tend to stress paradigms, pathdependencies and the temporal 'delay' in comparison to wider European transformations (see, for example, Giannakorou, 2005; Seixas and Albet, 2012) -can give only a partial understanding of the complex political and multi-scalar dimensions through which policies such as the PER are implemented in the long run, and their effects on local environments.In contrast, genealogical analyses are helpful in showing the contemporary presence of trends of continuity and change, and the non-linear trajectories of the diffusion of policy paradigms and tools; the interaction between a variety of economic, cultural and political factors; and the interaction of different actors across the local, national and supranational.This remains in line with recent critiques of the dominance of the 'paradigms' approach to policy studies and comparison (cf.Getimis, 2012;Tulumello, 2015).
In normative terms, understanding the trajectory of a supposedly 'old-fashioned' programme, such as the PER, can illuminate challenges that remain today remarkably similar for more 'advanced' policy schemes, such as the NGPH.
Finally, we want to list some key directions for further explorations that will complete and deepen our genealogic investigation.First, precisely because of the nature of the PER (which left the municipalities in charge of design and implementation), any evaluation of the programme's impact cannot be appropriately conducted without investigating its local dimension.For example, to what extent has the PER influenced the internal organisation of municipal administration, and the accumulation of resources in terms of knowledge and networks?In this respect, the framework of institutional learning and social innovation may provide useful paths (see Ferrão, 2011, pp.91-114).Secondly, the relation between housing, economic and financial policies seems to be a privileged point to understand housing policy more holistically (cf.Aalbers, 2016), for instance by exploring the way housing policies have often been 'concealed' as financial supports to home ownership while housing policy proper was being cut in the name of sound national finances.Finally, once one abandons the idea of benchmarking planning and housing systems and postulating a linear path of progress toward the best practices of more 'mature' systems, which alternative concepts can be mobilised to explain change, path dependencies and the persistence of long-standing features, and provide frameworks for comparison?It seems to us that, for instance, a theoretical framework of uneven development (cf.Hadjimichalis, 2011), much more than the comparison of developed/under-developed systems, is a fruitful approach to understand in depth the relations between planning and housing systems.

Figure 1 .
Figure 1.Distribution of Public Spending for Housing (above) and Public Spending in Rehousing (below) in Portugal,1987 -2011  (adapted from IHRU, 2015)   ) Some municipalities governed by the Communist Party initially resisted the implementation of the PER because they believed that housing policy was a state responsibility (former Secretary of State for Housing; Professor of Sociology and Urban Studies, interviews).Transactions of the Association of European Schools of Planning • 2 (2018) 32-46 Transactions of the Association of European Schools of Planning • 2 (2018) 32-46 Transactions of the Association of European Schools of Planning • 2 (2018) 32-46