2.1 Brazilian social housing

The lack of quality housing, combined with inefficient public policies, has made the supply of social housing in Brazil an ongoing problem since the mid-20th century to the present day (Moreira Cavalcante 2016; Dalla Vecchia & Medvedovski 2021). Brazil’s urbanisation process resulted in a huge housing deficit, estimated at around 5.8 million dwellings in 2019 (Fundação João Pinheiro 2021).

Since 2009, Brazil has constructed an unprecedented large number of social housing projects through the federal government’s PMCMV. In approximately 11 years, it has constructed 6 million housing units, with an investment of R$223.2 billion (Relatório de Avaliação, Programa Minha Casa Minha Vida 2020).¹ Between 2009 and 2019, the PMCMV covered families in three different income brackets, most notably low-income families earning up to three times the minimum wage.² The PMCMV aimed to boost the civil construction sector and reduce the country’s housing deficit by encouraging homeownership.

According to the rules of Caixa (the financing bank of the PMCMV), mortgage payments could be made up to 120 months, with monthly payments ranging from R$80.00 to R$270.00 depending on gross family income.³ The guarantee for the financing was the property itself, and most of the final value was financed by the government (Bortoli 2018). After paying all the installments, the beneficiary of the program became the owner, who had to observe some restrictions regarding the maintenance and alterations of the property during the financing period:

The execution of renovations and expansions in the property, which alter and/or compromise the installations and calculations of the original project, will result in the total loss of the warranty by the construction company, as well as if damages are found due to misuse or overloading of the permissible limits of the installations and original structures.⁴

The housing units offered by the PMCMV are characterised by a single standard typology: approximately 32–37 m² in area, of structural masonry construction, and a minimum accommodation comprising a living room, two bedrooms, bathroom, kitchen, and laundry area. This typology, for both single-story houses and apartment units is used repeatedly in practically all the different and complex Brazilian contexts. After a few years of occupancy, several changes⁵ by residents were observed, involving the removal of walls and the construction of new single-story rooms, mostly without technical assistance. Several studies carried out at national and local levels have demonstrated the extent of the problems faced by residents of social housing projects in the PMCMV (Amore et al. 2015; Villa et al. 2015). These standardised housing templates are disconnected from local social, economic, cultural, climatic, and environmental concerns (Formoso et al. 2011; Dalla Vecchia & Kolarevic 2020).

The procurement process for social housing has generally followed the rules and dynamics of the private real estate market, becoming primarily profit based and relegating housing quality. As a consequence, social housing estates are typically located in peripheral areas of the city (because of the low price of land), disconnected from the urban fabric, and without any consolidated or pre-established infrastructure. This increases its social and environmental vulnerability (United Nations 2013; John & Barros 2015; Biderman et al. 2019).

The dwellings produced have low functionality. The minimal architectural program disregards the needs of the different family profiles of the residents (Villa et al. 2017). The small size of rooms makes it difficult to install traditional basic furniture (Palermo 2013; Pereira 2017). POE studies (Villa et al. 2017; Parreira 2020) have found this is exacerbated by the transformations and (informal) remodeling of the housing over time, demonstrating the lack of projected flexibility and adaptability and of the forecast of extensions. By providing low-quality housing for the low-income population, the PMCMV lowers the resilience of the built environment and shows a low capacity to respond to the impacts and changes expected during their life cycle. As a result, recent research indicates the increase in the social, physical, and environmental vulnerability of these projects (Parreira 2020; Garrefa et al. 2021).

2.2 Resilience assessment in social housing

On a global scale, resilience is a critical urban development agenda, such as the New Urban Agenda of Habitat III (2017)⁶ and Sustainable Development Goals—AGENDA 2030 (United Nations 2015).⁷ Resilience is a driving force in combating the vulnerable state provided by the rapid growth of the urban population, and other problems caused by inadequate urbanisation (Meerow & Newell 2019; IPCC 2021). Resilience is here defined as the ability of the built environment to absorb, adapt, and transform itself to deal with the impacts imposed over time (Pickett et al. 2014; Hassler & Kohler 2014; Rodin 2015; Garcia & Vale 2017).

The impacts derive from major events (referred to here as major causes) which impose changes in the current situation, e.g. climate change, energy crises, political instabilities, etc. These may result in potential threats that arise both as interruptions—in short time horizons—and in the form of diffuse and slow threats—for longer periods (Garcia & Vale 2017; Heeren et al. 2015). These impact buildings as negative effects, derived from the interaction between the impact, the state of vulnerability of the system, and its adaptive capabilities. One way to measure this impact is to measure the level of disturbance caused by their negative effects.

Vulnerability is defined as the sensitivity of the system (built environment) to specific impacts, combined with the adaptive capacity of the population. Thus, it becomes relevant to assess incidental threats and specific vulnerabilities, which may lead to adapting actions that increase the resilience of an urban system (Davoudi et al. 2009). Adaptive capacity is closely related to the concept of adaptation, referring to the ability to adjust the behavior and characteristics of a system increasing its capacity to cope with external impacts and stresses (Maguire & Cartwright 2008; Brooks 2003).

Thus, the assessment of resilience implies a need for a deep analysis: analysing it both qualitatively and quantitatively, overcoming a simple predetermined checklist, and identifying its systemic behavior in the face of impacts (Garcia & Vale 2017; Garrefa et al. 2021; Villa et al. 2022). The resilience assessment proposed here is based on building performance evaluation and POE.

The utility of POE in achieving quality in architectural design has been demonstrated by studies in both the Brazilian (Ono et al. 2018) and international contexts (Roaf 2004; Leaman et al. 2010; Preiser et al. 2018; Li et al. 2018). The close relationship between the measurement of human behavior and the quality of domestic environment has been a way of raising the satisfaction rates of residents and improving performance. This involves the collection of assessments that include techniques for the physical perception of the built environment, as well as the interaction between this environment and user behavior (Cole 2005; Mallory-Hill et al. 2012; Villa et al. 2019). In this way, systematised POE in social housing can feed planning policies and practices that can address residents’ needs.

3.1 Research instruments

A Resilient House Evaluation Matrix was designed to identify the robustness of a system and, consequently, its adaptability, aiming to reduce vulnerabilities and expand adaptability. The matrix presents attributes and indicators for the promotion of resilient social housing. The attributes and indicators are based on: (1) theoretical and conceptual research performed on resilience assessment systems; (2) the work of Arup (a consultancy), which developed a methodology for analysing urban resilience named the City Resilience Framework (CRI);⁸ (3) the research group’s previous evaluation experiences; and (4) observations of more resilient housing practices.

The study considered the ‘attributes’ of resilience in the context of Brazilian social housing: environmental comfort, environmental adequacy, flexibility, accessibility, wellbeing, and engagement (Villa et al. 2022). Indicators, on the other hand, are the physical elements or social practices considered important to enable homes and their users to defend themselves from impacts, making them more resilient. The key attributes are summarised in Table 1.

Two objective POE tools were developed, and used by members of the research group since 2018. The Impact Assessment (IA) tool is a questionnaire that measures the level of disturbance to inhabitants caused by impacts on the houses, based on the residents’ awareness of the impacts, their frequency, and level of disturbance to residents, and the vulnerabilities are highlighted and the priorities for action established. Responses to questions regarding levels of disturbance were recorded as too much, too little, or not applicable. The option for a larger scale of discomfort measurement (e.g. from 0 to 5) was not chosen to avoid possible difficulties in understanding how to ‘measure’ one’s own perception. It was also clear that the respondents’ perception is guided by the negative effects that affect them, and not by the main causes. The main causes function more as an analysis tool for researchers than as a direct investigation of the residents’ perception.

In parallel, it is necessary to analyse the general performance of each attribute of resilience in view of an ideal configuration situation. The so-called ‘control cases’ define technical parameters to support the analysis of how near or how far is the case study in relation to an ideally resilient configuration.

3.2 Impact assessment (IA) tool

The home is subject to impacts of varied orders and, depending on its state of vulnerability at the time of occurrence, will absorb, adapt, and transform with greater or lesser ease, characterising its resilience. The term comprises the set of acute shocks and/or chronic stresses that threaten the lives, livelihoods, health, ecosystems, economies, cultures, services, and infrastructure of a society and built environment exposed, generating negative effects proportional to its state of vulnerability at a given time. Table 2 summarises the concepts involved in the idea of impact in the context of a social housing built environment.

The impacts on the urban environment and its systems are primarily caused by major global events. These challenges, which governments and society go through, are defined here as major causes. From the common impacts observed in social housing in previous research, six major causes of impacts for investigation were inferred (Table 3).

The IA tool was designed to collect residents’ perceived impacts, negative effects and level of disturbance, which explains the structure of the IA tool (Table 4). A questionnaire was provided to residents by means of a digital app.

The level of disturbance generated (too much, little, or not applicable—if nonexistent) was measured with 144 questions for each negative effect evaluated for the houses and another 140 for the apartment units.

3.3 Case studies: Sucesso Brasil and Oliva

The performance of the IA tool took place in two PMCMV developments (with residents from the lowest income bracket, i.e. household income 0–3 times the minimum wage) located in different sectors of the city of Uberlândia (MG). The city is characterised by a mild climate that alternates between a large hot and humid period (between October and April) and another cold and dry period (between May and September), with relative humidity in the range of 20% (Petrucci 2018; Bortoli 2018). Figure 1 locates the case studies in the city; Table 5 summarises the information that characterises the case studies: single-story houses at Successo Brasil Residential (RSB) and apartment units at Oliva Residential (RO).

3.4 Sampling and performance

There were 333 housing units from which a statistically representative sample of the population of 162 questionnaires were returned (80 in the RSB and 82 in the RO). The impact evaluation tool was administered between May 2019 and March 2020 (Table 6). This occurred nine years after the completion of the RSB and three to four years after RO. Residents were personally recruited by properly identified researchers carrying the IA questionnaire on their mobile phones, and the questions were asked and responses directly recorded in a virtual database.

5.1 Urban climate

When comparing RO and RSB, it was noted that low relative air humidity, deterioration of material, and need for an electric shower and air-conditioning were the topics that mainly disturbed residents (Table 8).

Regarding the results of negative effects for both sets, the main issues are related to the thermal discomfort caused by the dry season in the city, such as low relative humidity, hot temperatures and high level of dust, together with a deficiency in air and water tightness. The main issues of houses in the RSB are gaps between the roof tiles and the walls/beams that support them (Figure 5). Many residents reported that when heavy rains and/or winds occur, dust enters the dwelling. Expansion and contraction due to the high volume of air movement also causes cracking in tiles and ceiling, further compromising the building envelope.

In the RSB, the installation of solar water heating systems failed to waterproof around the pipes penetrating the roof (Bortoli 2018). As a consequence, mold occurs in 75% of cases in the RSB and in 27% of cases in the RO. The perception of the stuffy air in rooms due to the heat and poor air circulation was very concerning in 79% in the RSB and 39% in the RO, resulting in a threat to the respiratory health of residents.

In addition, thermal aspects are also generators of disturbances. Since natural ventilation is the main strategy for dealing with thermal discomfort in these dwellings, maintaining the flow of wind through the vents is essential in both sets. In the hot period, the windows are simultaneously a source of relief and discomfort, admitting wind and whatever else comes with it into the interior of the residences. To overcome this situation, the demand for fans and/or air-conditioning is high: 74% of respondents in the RSB and 61% in the RO reported this issue. Consequently, the cost of water and electricity in the hot period is a significant impact experienced by respondents, with greater intensity in the RSB (81%) and lower in the RO (60%) (Morais 2021). As a consequence, energy debts affected 22 out of 33 houses analysed in the RSB (Bortoli 2018).

These results show that the standard layout for a PMCMV housing unit, and its low quality of materials and construction, contributed little or nothing to the improvement of environmental conditions in relation to the elements of the urban climate.

5.2 Fuel poverty

Fuel poverty provokes a range of negative effects (Figure 2). The increases in the costs of energy and water bills are discomforts greatly perceived in both RO and RSB, as listed in Table 9.

It should be noted that in the RSB, houses have a solar water heating system that contributes significantly to the reduction of energy expenditures, which is approximately 80 kWh/month per household, or 20 kWh/person/month, considering an average of four inhabitants per household (Villa et al. 2017). The panels are always located to the north and west directions in order to capture as much energy as possible for the locality.

In both cases, the systems were found to heat the water sufficiently (Bortoli 2018). However, the volume of the solar water system (200 L) makes it necessary to ration the baths in families with more than four people or to use electricity for water heating which increases the costs of energy bills. This problem is more serious in the RSB as there are more occupants per household. A total of 74% of the RSB residents were very worried about their energy bills compared with 64% of RO residents.

5.3 Accelerated urbanisation

Accelerated urbanisation (Figure 2) refers to fast city growth, and whether the accompanying policies and urban design can keep pace (most of the time). This disconnect provokes a lack of quality in public areas, generating problems such as inadequate garbage disposal and lack of afforestation as visual pollution (Table 10).

The absence of a place for the disposal of garbage leads to complications. A total of 79% of the RSB residents have seen neighbors throwing garbage and debris onto the street or into empty lots, while 57% disposed of garbage and/or larger debris into empty spaces. In addition, 51% store building materials for future renovations in their backyards, often discovered and exposed to the weather (Figures 6 and 7).

As a consequence, the deposition of garbage in green areas generates a negative environmental impact in 86% of cases in the RSB, against 46% in the RO. As shown in Figure 6, the main green public area in the RSB has not yet been properly designed, becoming a focus for the deposition of garbage and debris (Figures 8 and 9). The lack of a designated recycling spot (a voluntary site for taking rubble, large objects, tree cuttings, and recyclable waste) in the vicinity of the RSB is reported by the residents as an aggravating factor, as the incidence of building alterations is high.

These and other questions make impaired aesthetics one of the most commonly perceived negative effects in the RSB, with 71% of respondents very concerned, while in the RO 37% are very concerned. In the RSB, 25% feel very concerned for not tending the public spaces of the neighborhood, while in the RO 32% feel uncomfortable not doing so, mainly due to the feeling of insecurity. This directly rebounds on important social costs for its residents, as community communication and engagement are key to achieve resilience. Collectively, residents might take charge in order to improve their public spaces; however, there is little to no sense of a real community bond, and the house problems surpass the collective issues, making it harder for changes to occur. This is deeply linked with how the development is implemented on-site.

5.4 PMCMV housing model

The rigid house layout prevents residents from easily adapting their spaces to their needs. From all the negative effects that come from the rigid housing model (Figure 3), the small size of the kitchen, poor quality of materials (problems with electrical, plumbing, or sewage installations), construction techniques (supporting walls that cannot be demolished/relocated), and lack of space for domestic activities are the most impactful items (Table 11).

The reduced room size and construction techniques makes it difficult for inhabitants to alter the layout. This is even more evident in the apartments because there is no possibility of expansion. It is interesting to note that the size of the kitchen is the source of the biggest complaints (82% in the RSB and 55% in the RO): most of the renovations start with the expansion of the back of the house to create a new kitchen (Figure 10).

In the RSB the possibility of expansion exists by using the yard, but this often proves to be tight and precarious (Figure 11).

In the RO, the common spaces do not meet this demand, with no spatial and design quality, directly affecting the environmental adaptation in the immediate surroundings, weakening the network of communication and engagement. This to a lack of sense of belonging. It reflects how the residents find meaning in living there, having a sense of enthusiasm, inspiration, and pride in relation to the neighborhood.

5.5 Incomplete application of the PMCMV

The aspirations of the PMCMV are to ensure the prosperity of the housing developments (e.g. the establishment of squares, leisure areas, family support programs, and a communication and support network for residents and local leaders). However, this has not been found in these cases. The lack of public facilities and social support fails to connect residents with public health, education. and labor services. But this is often not done. A wide range of negative effects occurs (Figure 4). Unused spaces eventually become source of illicit activities (Table 12).

When it comes to the relationship between residents and local leaders, the answer differs. Due to the physical configuration of the apartments, grouped into buildings, there is a closer relationship between the residents and the building manager. In the RSB, 90% are bothered by the lack of contact with existing leaders and activities. This lack of infrastructure, quality in common areas, and, consequently, lack of socialisation do not allow for the emergence of spaces that promote shared responsibility and trust. This comes as a major social cost since underprivileged communities tend to become more isolated. With this, the adaptability in the face of changes becomes even harder, with no ties between residents/neighbors and with agents of the community.

5.6 Socioeconomic factors

Families living in PMCMV housing often have low income or come from socially and economically marginalised communities (Figure 4). This, associated with a house of poor quality, ineffective guidance in the owner’s manual, and lack of a relationship with leaders and policies, makes it even harder for residents to adapt in a positive way (Table 13).

The biggest problem found in these low-income families is the difficulty in saving money, which affects 74% of the residents of the RSB and 65% of the RO. This factor is aggravated by deficiencies and spatial problems found in the homes, which force these families to carry out urgent renovations and repairs, compromising the family budget, and, consequently, impacting the family’s dynamics and quality of life.

In the RSB, the perceived disturbance with renovations is greater, with 67% not undertaking necessary renovations and 54% interrupting renovations due to lack of income. Such disturbances are aggravated by the lack of adequate technical assistance, and generate further difficulties, compromising the ability to flourish.

Inhabitants’ greatest perceived disturbances are the consequences of a poor-quality project, which increasingly highlights the problems of climate, weather, urbanisation, security, space, and how this affects the social sphere of these communities. The impact evaluation tool identifies these issues and creates a shared narrative, where the analysis of researchers/technicians and the residents’ vision find a common place to start changes.