Changing land-use metrics in mass housing: Türkiye case study

1.1 HISTORICAL DEVELOPMENT OF MASS HOUSING

Mass housing have evolved continuously from the 19th century to the present, reflecting changing spatial organisations, urban fabric relations and design philosophies. The Industrial Revolution in Britain triggered the first systematically organised developments, where worker settlements such as Saltaire (1851–71) near Shipley, West Yorkshire, and Bournville (some houses by 1879, model village began 1893), near Birmingham, used rigid grid plans, closed-block configurations and centrally located green space to address functional monotony (Benevolo 1980; Creese 1966). These early models were characterised by high-density construction and minimal green space.

A fundamental reorientation occurred with Ebenezer Howard’s ‘Garden City’ concept (Howard 1902). Projects in Britain, such as Letchworth (1903) and Welwyn (1920), both Hertfordshire, introduced radial layouts and low-density configurations, prompting a reassessment of open space-to-built area ratios. Concurrently, continental examples such as Hellerau (1909) in Dresden, Germany, and Betondorp (1923) in Amsterdam, the Netherlands, integrated geometric plans with social amenities and green corridors (Bullock 1988).

The Modernist movement introduced the ‘open block’ system to optimise natural light and ventilation. In Germany, Ernst May’s Römerstadt (1927–28) and Praunheim (1929) complexes in Frankfurt, along with Bruno Taut’s Hufeisensiedlung (1925–30) in Berlin, exemplified the integration of elongated housing blocks and expansive green strips (Henderson 2013; Scarpa 1990). The Congrès Internationaux d’Architecture Moderne (CIAM—International Congresses of Modern Architecture) further popularised functional segregation, manifested in J. J. P. Oud’s Kiefhoek (1925–29), in Rotterdam, the Netherlands, which featured standardised units and systematically arranged open spaces (Taverne 1996).

Post-Second World War housing shortages led to large-scale complexes. Le Corbusier’s ‘Radiant City’ advanced high-rise residential blocks within park landscapes to balance density and open space (Fishman 1977). In Britain, London’s Alton Estate (1955–59), in Roehampton, combined slab-and-point block typologies with generous playgrounds and social amenities (Glendinning & Muthesius 1994). In the United States, Robert Moses established new standards for large-scale housing through superblocks and centralised parklands in projects such as Stuyvesant Town (1947) and Peter Cooper Village (1947)—‘StuyTown’—in Manhattan, New York City (Caro 1974).

The 1960s introduced ‘megastructures’ and novel spatial configurations. In Canada, Moshe Safdie’s Habitat 67 (1967) in Montreal, Quebec, utilised modular arrangements to offer multi-tiered open spaces, while in Britain, Denys Lasdun’s Ziggurats (1962–68) at the University of East Anglia, Norwich, featured topographically responsive terraced student housing (Safdie 1970; Harwood 2015). In Japan, Kenzo Tange’s Tsukiji Redevelopment (1963) in Tokyo manifested the Metabolist ethos through modular growth and infrastructure integration (Lin 2010).

Social shortcomings in the 1970s, illustrated by the demolition in the US of Pruitt-Igoe (1972), in St Louis, Missouri, prompted a reassessment of Modernism (Newman 1972). Defensible space theory and user participation became central, as seen in Herman Hertzberger’s Diagoon Houses (1971) in Delft, the Netherlands, and Ralph Erskine’s Byker Wall (1969–82), in Newcastle-upon-Tyne, UK, which emphasised social interaction areas and topography (Erskine 1973; Hertzberger 1991). During this period, Martin & March (1972) provided a seminal reference for urban morphology, systematically analysing the relationship between FAR and BCR to demonstrate that density does not necessitate uniform form.

The 1990s New Urbanism movement, seen in projects such as Seaside (1981) and Celebration (1994), both in Florida, US, reinvigorated traditional street-centred organisation and mixed-use principles (Katz 1994). Simultaneously, the Compact City Theory, advanced by Newman & Kenworthy (1989), advocated for intensified development (high FAR) and transit-oriented systems to reduce car dependency. They argued that increasing FAR enhances compactness, provided BCR is balanced to protect open spaces. In China, the ‘Danwei’ system integrated residential and work functions within self-contained compounds featuring central green spaces (Bray 2005).

In the 21st century, sustainability became a primary criterion. Bjarke Ingels Group’s 8 House (2009), in Copenhagen, Denmark, Freiburg’s Vauban (1998–2006), in Germany, and MVRDV’s Parkrand (2006), in Amsterdam, the Netherlands, exemplify energy-conscious designs with vertical gardens and green corridors (Ingels 2009; Nobis 2003; MVRDV 2005). Recent research further explores the effects of density and building orientation on public space shading, surface temperatures, and heat gain (Asfour et al. 2023; Chung & Yoon 2023). Glaeser (2011) argues that a high FAR promotes economic efficiency and innovation, while Bertaud (2018) and Angel (2012) emphasise that urban productivity is determined by accessibility and the alignment of land-use metrics with market signals.

Historical benchmarks for land-use metrics include the 1926 US National Recreation Association (NRA) target of 40.5 m² per capita. Today, the World Health Organization (WHO) recommends a minimum of 9 m² of green space per person for basic health, with an ‘ideal’ 50 m² for a high quality of life. Success in meeting these targets varies significantly based on development models.

Cities following planned growth strategies often exceed these standards. Singapore leads globally with 66 m² per capita (Hassan & Jombach 2026), using a ‘City in a Garden’ vision that employs vertical greening and sky gardens to compensate for a high FAR (Yuen 1996). Curitiba, in Paraná, Brazil, reaches the ideal 52 m², while Sydney in New South Wales, Australia, maintains 27 m², well above WHO limits.

Conversely, rapid urbanisation in developing nations often results in green space provision lagging behind population growth. Tehran, the capital of Iran, provides only 7.2 m², falling below the WHO minimum. Extremely dense environments such as Hong Kong (3.0 m²), in China, and Mumbai (1.2 m²), in India, face critical shortages (Hassan & Jombach 2026). In Hong Kong, high BCR levels lead to shadowing and blocked wind corridors, which degrade the ecological quality of limited existing green space (Jim 2004).

In China’s megacities such as Shenzhen and Guangzhou, a ‘green inequality’ is observed where high-density gated communities contain private luxury landscapes that do not contribute to the city’s general green continuity (Wu et al. 2024). In Dhaka, Bangladesh, the situation is more severe: as built-up areas rose from 59% to 82% between 1989 and 2020, qualified green space plummeted from 17% to just 2% (Nawar et al. 2022). These disparities underscore the global challenge of maintaining balanced land-use metrics amidst increasing urban density.

1.2 MASS HOUSING IN TÜRKIYE

The period between 1923 and 1961 represents the foundational era of housing policies in Türkiye. Although the modern concept of mass housing was not yet established, the first initiatives for low-income groups were implemented through state-supported cooperative models. Cooperative movements beginning in the 1930s aimed to facilitate access to affordable housing for civil servants and worker families (Keleş 2008). The first legal standard for green spaces was introduced by the Municipal and Building Roads Law No. 2290 in 1933; under Article 4(b), a per capita open and green space (groves, meadows, lakes and playgrounds) requirement of 4 m² was proposed, targeting a 6% urban green space ratio.

The period 1961–80 represents the beginning of the planned development era in Türkiye, characterised by rapid industrialisation and significant rural-to-urban migration challenges. The absence of a structured urbanisation strategy during this period triggered the squatter housing problem (Tekeli 2010). In 1967, the term ‘mass housing’ was officially introduced in the Second Five-Year Development Plan, shifting the state’s focus toward social housing production for low-income groups. The establishment of the General Directorate of Land Office in 1969 was a crucial step in organising mass housing production (Ulusoy 2020). Zoning Law No. 6785 (1956) stipulated that green space could not be less than 7 m² per capita, though it remained unclear how this standard should be differentiated across various settlement scales.

Post-1980, Türkiye’s housing sector underwent a neoliberal transformation, where the state’s role was redefined within a free-market economy. The Mass Housing Administration (TOKİ), established in 1984, became a central actor in social housing production for low- and middle-income groups (Tekeli 2010). Zoning Law No. 3194 (1985) maintained the 7 m² per capita green space requirement for city centres, while increasing it to 14 m² for planning areas outside municipal and adjacent boundaries.

A defining characteristic of mass housing projects in the 1970s and 1980s was the priority given to socialisation spaces. Housing complexes were designed as spatial organisations supporting social life, with playgrounds, sports facilities, cultural centres and green spaces functioning as vital interaction points. These areas fostered neighborhood relations and social cohesion, serving as essential elements for addressing urban cultural issues and enhancing quality of life. Density policies during this period integrated social and environmental factors into FAR and BCR criteria (Ulusoy 2020). Building distances were maintained much wider than current standards to protect privacy and provide adequate ventilation and sunlight exposure.

From the late 1990s, neoliberal policies shifted housing production toward security-oriented and enclosed ‘gated communities’. In these models, public socialisation spaces were replaced by restricted common areas accessible only to residents, leading to weakened social solidarity and increased individualisation in urban life (Tekeli 2010). Economic priorities disrupted the solid–void balance, dramatically increasing construction ratios and minimising open spaces. As noted by Harper (2013), reduced distances between tall buildings negatively impact urban life through blocked solar radiation and wind barriers; nevertheless, modern projects frequently utilise maximum FAR and BCR ratios while reducing distances to minimum levels.

A primary driver of this transformation is the inclusion of FAR exemptions within the regulations, which created the excluded area loopholes. These exemptions had long been defined by various municipal zoning regulations before their national formalisation through Article 22 of the Planned Areas Zoning Regulation (2017). These loopholes, which include roof gardens, places of worship, parking, mechanical rooms and balconies, can constitute approximately 30% of the total floor area, leading to built environments that significantly exceed the construction densities originally prescribed in urban development plans.

Holistic studies on urban density and standards in Türkiye are remarkably limited. Ersoy (2015) documents these across multiple scales, noting dramatic disparities such as the 0.07 FAR in Balgat compared with the 6.26–7.31 FAR values in Mesa and Yüksel Sitesi. Green space adequacy data from the early 1970s reveals significant gaps: Antalya (4.32 m²), Karadeniz Ereğlisi (3.27 m²), Elazığ (1.87 m²), Kars and Aydın (1.82 m²), Antakya (1.56 m²), Balıkesir (1.37 m²), Erzincan (1.28 m²), Mersin (1.17 m²), Denizli (1.01 m²), Kütahya (0.93 m²), and Edirne (0.70 m²) (Gürel 1972). Çetiner (1972) also highlights significant disparities, where leaders such as Sivas (16.00 m²) and Balıkesir (9.03 m²) contrasted with severe deficiencies in İzmir (2.60 m²) and Konya (3.20 m²) (Ersoy 2015).

The literature lacks macro-scale holistic studies beyond Çetiner (1972), Gürel (1972) and Ersoy (2015). Most other research focuses on city-level analyses, such as: Ankara, 2.3 m²; Istanbul, 2.1 m²; İzmir, 2.8 m² (Oruç 1992); Edirne, 10.5 m² (Uysal 1995); Kayseri, 5.43 m² (Öztürk 2004); Eskişehir, 4.6 m² (Aksoylu et al. 2005); Kahramanmaraş, 1.4 m² (actual versus 7.1 m² planned) (Doygun & İlter 2007); Niğde, 4.09 m² (Olgun & Yılmaz 2019); and Çankırı, 4 m² (Koçan & İbiş 2020). While these findings illustrate recreational adequacy at the city level, land-use metrics within mass housing areas have not yet been academically studied in the Turkish context.

In summary, housing production models and spatial metrics in Türkiye are dynamic and redefined by social needs (Tekeli 2010). Rather than static comparisons, tracing the evolution of spatial forms and the cumulative impacts of these metrics over time provides a more rational framework for understanding urban development.

1.3 MASS HOUSING IN KOCAELI-İZMIT

Strategically located on the transportation axis between Istanbul and Ankara, İzmit serves as a critical junction connecting Europe and Asia. Its natural harbour and proximity to Istanbul have established the city as a leading industrial centre with rapid urbanisation (KBB 2009). Although İzmit represents a significant case for Türkiye’s industrialisation and urbanisation, there is insufficient research regarding the spatial development process of the city and the specific role, position, status and effect of mass housing areas within this process (Figure 1).

Figure 1:

Historically, the city initially developed linearly along a narrow coastal band. However, the 1999 Marmara Earthquake became a definitive breaking point; growth shifted from the east–west coastal axis toward the more resilient ground structures in the north and south. Subsequent industrial zoning and increasing population led to a macroform characterised by dense build-up in the north-west and eastern sectors by 2020 (see Figure S1 in the supplemental data online).

The evolution of İzmit’s mass housing has been dictated by its industrial character and disaster history. The earliest examples emerged in the 1950s as worker housing for major facilities such as TUPRAS (1961) and SEKA (1962), utilising functionalist grid plans on reclaimed land (Gezici et al. 2010). Kemal Ahmet Aru’s 1959 plan institutionalised this development by relocating the D-100 highway and zoning worker housing in proximity to industrial areas (Aktaş et al. 2023).

In the 1980s, inadequacies in official policies led to spontaneous settlements in areas such as Erenler and Kozluk, while institutional housing remained restricted to two- to three-storey cooperative blocks. By the 1990s, the private sector introduced gated communities with integrated security and social facilities (Bal 2015).

The 1999 earthquake caused heavy damage to 66,441 houses and displaced approximately 200,000 people. This crisis necessitated a transition from temporary prefabricated settlements to permanent, earthquake-resistant projects led by TOKİ and Kentkonut (TOKİ 2015). Since 2010, urban transformation has further accelerated the renewal of the old building stock (KçşİM 2020).

Regional studies on green space reveal significant deficits, despite the high value residents place on housing–green area relationships (Çepni & Kutluca 2024). Sönmez (2022) identified a per capita green space of 5.89 m² for Kocaeli and 5.65 m² for İzmit district. Neighborhood-scale data highlight disparities between Akarca (5.15 m²) and Cedit (0.04 m²). Furthermore, district-wide land use for parks remains limited: urban/district parks cover 0.91%; neighborhood parks, 5.63%; children’s playgrounds, 0.72%; and sports areas, 0.10% (Ihlamur 2019).

In summary, İzmit reflects a broader trend where per capita green space remains significantly below the targeted 10 m² national metric, underscoring the challenges of balancing industrial density with adequate open space.

2.1 RESEARCH DESIGN AND STUDY AREA

The study aims to quantitatively assess changes in land-use strategies and the associated changes in building density in both recently developed and urban transformation areas. To this end, public mass housing projects built during a defined period in a major industrial city, considered representative of Turkish averages with respect to both political context and lifestyle, were selected as the sample (KBB 2009). A descriptive mixed-methods case study methodology, integrating spatial analysis with expert evaluations, is employed to investigate the relationship between building density and land-use in mass housing projects. The research encompasses a comprehensive inventory of eight publicly developed mass housing projects in Kocaeli-İzmit constructed after 1990, providing a longitudinal analysis of design evolution in a specific regional context. The adoption of a comparative case study approach, rather than an inferential statistical method, was chosen because of the fixed number of comparable public projects, which allowed for detailed observation of temporal changes to be made.

2.2 QUANTITATIVE DATA COLLECTION AND COMPARATIVE ANALYSIS

The initial stage comprises a spatial analysis based on multifaceted data collection including site plans (provided by İzmit Municipality), base maps, satellite imagery (provided by Kocaeli Metropolitan Municipality) and attribute data (the authors’ own on-site measurements and office calculations on computer-aided design (CAD) platforms). The key morphological parameters calculated were: realised floor area ratio (FAR), defined as the ratio of the actual total construction area to the project area; realised building coverage ratio (BCR), the ratio of the actual area occupied by housing blocks on the ground to the total project area; and the building proximity index (BPI), defined as the ratio of the distances between building blocks to the height of the block.

Beginning with attribute data, comparison parameters for the physical design of social housing were generated through the following processes.

FAR and BCR were calculated as the ratio of the total building area and the building footprint area, respectively, to the total project area:

The areas of active green spaces, playgrounds, sports facilities and car parks were measured and the amount of space allocated per dwelling calculated.

For BPI, first, the average distance of each block to the nearest three to five neighbouring blocks was divided by the block’s height, and a proximity values (P) were calculated for each block separately (equation 3). The BPI was then obtained by averaging all proximity values within the project area (equation 4). This index is used as a tool to evaluate construction density and spatial spaciousness in urban areas.

This study examines eight mass housing projects constructed after 1990 in the İzmit district of Kocaeli (Figures 2 and 3; for detailed site plans, see Figure S2 in the supplemental data online). Projects developed only by the central or local governments were included to avoid limitations in comparability.

Figure 2:

Figure 3:

Amenity areas, specifically active green areas, playgrounds, sports areas and parking, were measured and normalised by calculating the amount of space per dwelling unit to eliminate project size bias. The analysis focused on longitudinal observation and direct comparison of these normalised metrics across the project timeline to identify temporal trends and critical shifts in land-use allocation.

2.3 QUALITATIVE DATA COLLECTION AND THEMATIC ANALYSIS

The second stage involves face to face semi-structured expert interviews (on spatial usage density, impacts of legislations and policies, physical and environmental design criteria, social amenity priorities, and quality of life) to provide a context for the quantitative findings and to assess the reasons for observed differences (Figure 4). Structured interviews were conducted with a sample of 15 local professionals. For the questions asked, see Table S1 in the supplemental data online. The experts consisted of five architects, four urban planners, four surveying engineers and two civil engineers from both public and private sectors. The qualitative data obtained from the interviews were subjected to thematic analysis to systematically derive causal explanations for the measured changes, which involved transcribing the interviews and inductively coding the content to group responses into overarching themes concerning policy, economics and design priorities.

Figure 4:

3.1 LONGITUDINAL SHIFTS IN CONSTRUCTION DENSITY

This study analysed data from mass housing projects to observe the building land-use relationship and examine temporal change in amenity areas. The construction area measures (Table 1) were normalised by calculating their proportions relative to amenity-area measures, enabling comparative analyses based on standardised metrics.

This study designates the actual building coverage ratio (A.BCR) and actual floor area ratio (A.FAR) as primary density indicators, since these realised values better reflect the final land-use than initial development plans. Chronological analysis of the data reveals a clear shift in density (Table 2 and Figure 5). The early projects, specifically those constructed between 1990 and 2000, displayed relatively low-density figures, with A.FAR values stabilising around 0.60. In contrast, the post-2008 era marks a substantial vertical and horizontal densification. Projects constructed after this period consistently show A.FAR values approaching or exceeding 2.00, representing an increase of over 300% compared with the 1990 benchmark project. By definition, an increase in BCR and FAR ratios is expected to correspond to a decrease in the area of amenities. In this regard, observed changes in green areas, parking areas, playgrounds and sports facilities further confirm the findings of the density analysis.

Figure 5:

An examination of the Yildiz Housing project (2013) highlights a specific anomaly in this trend. This project records the highest A.FAR of 3.24 and A.BCR of 0.19 among all sampled areas (see Figure S3 in the supplemental data online). Although this development targets a middle-income demographic and uses underground parking, the analysis reveals that the surface void created by burying the parking infrastructure was not used for open communal space. Instead, it was transformed into additional building mass. The dramatic increase in both FAR and BCR values following the 2000s confirms a consistent pattern of aggressive land use, with the primary objective appearing to be the maximisation of construction volume.

3.2 DEGRADATION OF AMENITY PROVISION

An increase in building density is directly associated with a corresponding decrease in open spaces and amenity areas per dwelling unit. Active green areas per housing unit show the largest decline over the sample time frame. The Yahya Kaptan project (1990) provided an exceptional 102 m² of active green area per dwelling unit. In contrast, projects constructed after 2008 provide an average of only 20 and 26 m² per unit. Recent figures on active green spaces are significantly below the WHO minimum standard of 9 m² per person. Considering an average household size of 3.6 persons, the current provision of 20–26 m² per unit equates to only 5.5–7.2 m² per capita, falling well short of WHO recommendations. This trend indicates a reduction in available space per housing unit for social interaction and public green spaces, confirming the quantitative degradation of the urban environment.

Furthermore, the BPI, which serves as a reference for distances between structures, indicates that living comfort has been compromised. The decreasing trend in BPI confirms that taller, denser buildings are increasingly preferred over spacious layouts. The two earliest projects (1990 and 2000) boast the highest BPI values, indicating generous spacing between blocks relative to their heights. Conversely, later projects such as İzmitkent-4 record extremely low index values, signifying a severe narrowing of distance between building blocks. This reduction negatively impacts usage comfort as well as access to sunlight and natural ventilation. Even when the FAR is not exceptionally high, a low BPI suggests a design preference for low-rise, densely packed blocks, resulting in a negative solid–void balance.

With respect to additional amenity provisions, although playground and sports area ratios per housing unit remain relatively stable over time, this stability only occurs alongside a significant reduction in open green areas. For example, the 1990 project stands out for its provision of 6.85 m² of sports area per unit, a result of its larger project area and complex urban design. In contrast, certain projects from 2010 and 2011 do not include any sports facilities. The 2013 project addresses parking adequacy through underground solutions; however, the overall trend indicates that amenity spaces are being reduced to allow for greater building mass.

3.3 EXPERT EVALUATION: THEMATIC ANALYSIS OF DENSITY DRIVERS

To interpret the quantitative findings and explain the underlying causal mechanisms contributing to the observed increase in density increase, the results were contextualised through a thematic analysis of interviews with 15 sectoral experts. Participants were identified based on their mass housing projects experiences, professional tenure and functional roles. To ensure a balanced representation across the public sector, private industry and contracting entities, senior practitioners were selected from the disciplines of architecture, urban planning and geomatics engineering. For details about the qualifications of these experts, see Table S2 in the supplemental data online.

The qualitative data yielded three primary themes that explain the transformation of the built environment.