Building age diversity and retail patterns: A morphometric exploration of Jacobs’ neglected condition through the Temporal Fingerprint Matrix

Abstract

Chinese

Jane Jacobs established diversity as the cornerstone of urban vitality, arguing that four conditions – mixed primary uses, small blocks, buildings varying in age and concentration of people – generate the exuberant diversity essential for vibrant cities. Recognizing retail patterns as indicators of urban vitality, urban studies has extensively tested three conditions, while building age diversity remains the least examined. The few existing studies have produced conflicting findings due to limitations in how building age characteristics are measured. To address these gaps, we introduce the Temporal Fingerprint Matrix, a novel analytical framework integrating building age and age diversity to distinguish incremental from instant development patterns. Aligned with Jacobs’ theoretical framework, we hypothesize that incrementally developed areas, characterized by diverse building ages and substantial older building stock, demonstrate higher retail density and diversity compared to instantly developed areas with uniform and predominantly newer buildings. Analysing approximately 600,000 buildings across Amsterdam, Rotterdam and Den Haag, we employ hierarchical clustering and identify distinct development types. Statistical analysis examines whether these types systematically influence retail density and diversity. Results reveal seven distinct development types with significantly different retail patterns and confirm our hypothesis: incrementally developed areas consistently outperform instantly developed areas in retail density and diversity. This study fills a critical knowledge gap by providing the first systematic test of Jacobs’ most neglected diversity condition and introduces a replicable morphometric method for identifying temporal development patterns. The findings inform time-conscious planning strategies for enhancing retail density and diversity through balanced preservation-development approaches that cultivate age-diverse built environments.

Keywords

building age diversity incremental development Jane Jacobs morphometrics retail density and diversity temporal development patterns

Introduction

Jane Jacobs’ (1961) The Death and Life of Great American Cities remains a foundational reference in urban studies, establishing diversity as the cornerstone of urban vitality and shaping how the field values mixed uses, density, accessibility and walkability (Delclòs-Alió and Miralles-Guasch, 2018; Gehl, 1987; Gómez-Varo et al., 2022; Hirt and Zahm, 2012; Klemek, 2007; Southworth and Ben-Joseph, 1997; Talen, 2010). Central to Jacobs’ framework is the recognition that retail density and diversity reliably indicate broader urban vitality. She argued that:

wherever we find a city district with an exuberant variety and plenty in its commerce, we are apt to find that it contains a good many other kinds of diversity also, including variety of cultural opportunities, variety of scenes, and a great variety in its population and other users. (Jacobs, 1961: 148)

Retail patterns thus serve as an observable proxy for the underlying conditions that generate urban diversity and vitality, as contemporary research confirms that these services generate essential social and cultural capital beyond basic material needs (Meltzer and Capperis, 2017). Jacobs systematized these conditions as four generators: mixed primary uses ensuring overlapping activities; small blocks increasing pedestrian movement; buildings varying in age and condition creating diverse economic opportunities; and dense concentrations of people providing critical mass (Jacobs, 1961: 143–221).

Despite extensive empirical engagement with Jacobs’ framework, research attention has been markedly uneven across these four conditions. Mixed primary uses have been widely operationalized yet yield inconsistent findings. Some studies demonstrate positive associations with urban vitality (Long and Huang, 2019; Wu et al., 2018; Yue et al., 2017; Zeng et al., 2018), while others report no significant relationships (De Nadai et al., 2016; Sung et al., 2015). Small blocks have attracted substantial empirical support, with studies consistently demonstrating positive associations with pedestrian walking behaviour (Ewing and Cervero, 2010), mobile phone activity (De Nadai et al., 2016) and commercial vitality (Long and Huang, 2019). Concentration represents the most extensively studied variable (Huang et al., 2023), revealing positive associations with population density (Zeng et al., 2018), GPS-based pedestrian concentration (Wu et al., 2018) and built environment density (Montgomery, 1998; Ye et al., 2018). In contrast, buildings varying in age and condition have attracted disproportionately limited attention, despite occupying a special position within Jacobs’ theoretical framework (Grant, 2017; King, 2013; Powe et al., 2016).

Addressing this knowledge gap, we examine how building age characteristics influence retail patterns, which is a relationship that remains an empirically untested part of Jacobs’ theoretical framework. Following Jacobs’ recognition that retail patterns serve as a proxy for broader urban vitality rather than a direct measure of it, we ask whether areas with distinct age characteristics exhibit significantly different retail density and diversity patterns. The following sections elaborate the theoretical foundations of this neglected condition (section ‘The value of older buildings and building age diversity’) and introduce our integrated analytical approach that we name the ‘Temporal Fingerprint Matrix’.

The value of older buildings and building age diversity

The theoretical foundation for this neglected condition rests on two complementary dimensions: an economic logic concerning the value of older buildings, and a procedural logic concerning how urban environments develop over time. While previous research has engaged with these dimensions separately, we argue that Jacobs’ third condition can only be understood through their integration.

The economic dimension is well established in theory. New buildings demand higher rents to recover construction costs, limiting occupancy to established operations like chain stores and banks. Conversely, older buildings offer affordable spaces where diverse businesses like neighbourhood bars, small restaurants and speciality stores can operate successfully (Jacobs, 1961; Sung et al., 2015). Providing affordable space for this ‘experience economy’ is increasingly vital today, as these independent venues help city centres maintain resilience against retail decline (White et al., 2023). This affordability also proves crucial for innovation, as experimental ventures require low-cost environments for trial and error – a principle reflecting Jacobs’ insight that new ideas must use old buildings (Bettencourt, 2021). Consequently, the presence of older buildings establishes the economic foundation for a diverse commercial ecosystem.

However, empirical evidence for this economic logic remains inconclusive. Studies on older buildings yield mixed outcomes: some show positive associations with pedestrian activity (Sung et al., 2015) and commercial density (Huang et al., 2023), while others find no effects on urban vitality (De Nadai et al., 2016). Moreover, legitimate counter-arguments challenge the affordability premise directly, arguing that preservation may constrain supply and increase property values, and eventually lead to commercial gentrification rather than supporting affordability (Bantman-Masum, 2020; Been et al., 2014; Glaeser, 2012).

These limitations demand attention to the complementary dimension. Jacobs did not simply advocate for old buildings but for environments where ordinary, low-value older buildings coexist with buildings of different ages and conditions (Powe et al., 2016). This shifts the focus from building age alone to building age diversity as a distinct variable. Yet studies examining age diversity in isolation present parallel problems: some studies demonstrate positive correlations between age diversity and social or economic vitality (King, 2013; Powe et al., 2016), while others report no relationships (Sung et al., 2015). Although recent computational approaches confirm that building age patterns influence urban vitality across multiple contexts (Delclòs-Alió and Miralles-Guasch, 2018; Gómez-Varo et al., 2022; Yoshimura et al., 2022), they maintain this problematic separation of age-related variables, namely age diversity and building age, and prevent a nuanced understanding of their combined effects.

These conflicting results reflect a fundamental analytical limitation: existing research treats building age and age diversity as independent variables, though Jacobs’ third condition inherently requires their integration. Unlike mixed uses, short blocks and concentration which can be achieved through planning, buildings varying in age and condition cannot be created instantly but emerge through temporal development processes (Grant, 2017; King, 2013). Recognizing this, we propose the Temporal Fingerprint Matrix, an integrated conceptual framework that brings these two dimensions together to characterize distinct temporal conditions of urban form and resolve the inconsistencies that arise from examining them in isolation (Figure 1).

Figure 1.

The Temporal Fingerprint Matrix.

The Temporal Fingerprint Matrix

The Temporal Fingerprint Matrix provides a conceptual framework for operationalizing Jacobs’ third condition by integrating building age and age diversity as complementary indicators of urban development processes. Our dual matrix places building age on the horizontal axis and building age diversity on the vertical axis, creating a bivariate analytical space representing four distinct urban development archetypes (Figure 1).

Incremental development (top right) combines high age diversity with predominantly older buildings, representing Jacobs’ ideal urban condition. These areas have accumulated through generations, where ongoing cycles of replacement and infill create the temporal stratification.

Instant development (bottom left) represents the contrasting development pattern, characterized by low diversity with predominantly newer buildings. This pattern reflects a total design approach executed within short timeframes, corresponding to Jacobs’ critique of large-scale construction projects that fail to support diverse activities due to their temporal uniformity.

Preserved development (bottom right) combines low diversity with predominantly older buildings, characterizing historical areas that remain frozen in time with building stock from consistent periods. While these areas possess older structures, they lack age diversity.

Recent incremental development (top left) depicts high diversity with a predominance of newer buildings, suggesting areas that maintain incremental character despite their relative youth. These areas incorporate temporal variation through mixed-age construction or selective preservation during redevelopment, demonstrating that incremental processes can operate even within contemporary development contexts.

These four archetypes provide conceptual clarity, though empirical reality may reveal transitional and hybrid conditions at quadrant boundaries.

In this article, we operationalize building age and age diversity through the Temporal Fingerprint Matrix to identify distinct development patterns and examine their influence on retail density and diversity, which Jacobs recognized as indicators of broader urban vitality. We hypothesize that incrementally developed areas, characterized by diverse building ages and substantial older building stock, will demonstrate higher retail density and diversity compared to instantly developed areas with uniform and predominantly newer buildings. To test this hypothesis, we employ hierarchical clustering to identify distinct temporal development types across approximately 600,000 buildings in Amsterdam, Rotterdam and Den Haag. Then, we examine whether these development types show statistically significant differences in retail patterns.

The remainder of this article proceeds as follows. We detail our methodology in the ‘Methodology’ section, including dataset development, hierarchical clustering and statistical analysis. After presenting the results in the ‘Results’ section, we discuss findings, contributions and future research directions in the ‘Discussion’ section and present concluding remarks in the final section.

Methodology

Case studies

The Netherlands provides exceptional analytical conditions for examining building age characteristics and their effects on retail patterns. Previous research in highly preserved European contexts has fallen short; for example, Italian cities showed no significant relationships between aged buildings and urban vitality due to the overwhelming prevalence of historic structures constraining analytical variation (De Nadai et al., 2016). Dutch urban environments exhibit remarkable temporal heterogeneity, combining well-preserved historical districts with wide-ranging contemporary developments. This variability reflects the Netherlands’ distinctive urbanization trajectory, bringing together incrementally developed and preserved areas alongside large-scale instantaneous development (Rutte and Abrahamse, 2016; Van der Valk, 2015).

We select the big three cities presenting the most complex urban development conditions. Amsterdam represents complex preservation-development dynamics, combining organically evolved historical districts with contemporary planned constructions (Savini et al., 2016). Rotterdam constitutes an exceptional post-war reconstruction case following Second World War bombing that eliminated substantial portions of the city core, creating space for large-scale instant development (Mccarthy, 1999). Den Haag maintains its incrementally developed historical centre while experiencing intensive large-scale peripheral development (Nabielek et al., 2013). To capture wide-ranging metropolitan coverage, we define study areas using 15-km radii from each city centre.

Dataset development

Our analysis relies on two primary data sources (Figure 2). First, we utilize the Dutch 3D-BAG dataset, which provides building-level construction dates for approximately 600,000 buildings across our three cities. Following Wagenaar’s (2016) periodization of Dutch urban planning history, we organize buildings into eight temporal categories corresponding to major planning periods: pre-industrial urban development characterized by compact traditional city cores (1600–1830); early industrialization marked by growth and densification of Dutch cities (1831–1870); planning emergence with the rise of Dutch planning tradition and the first housing law of 1901 (1871–1915); modernist expansion representing the heyday of urban planning and comprehensive extension plans (1916–1945); post-war reconstruction focused on new housing provision and urban renewal (1946–1970); suburban expansion characterized by infrastructural highway development and city expansion (1971–1990); a differentiation period breaking visual and functional monotony through diverse planning ideals (1991–2010); and liberal planning emphasizing market-orientated and open-ended urban development approaches (2011–2025). Data-quality verification involved removing 791 buildings coded with a construction year of 1005, representing placeholder values for uncertain historical dating.

Figure 2.

Methodological framework of the study.

Second, we source retail establishment data from OpenStreetMap’s Points of Interest database, extracting businesses using specific OSM tags and grouping them according to the corresponding NACE division groups (statistical classification of economic activities in the EU) to ensure replicability across European contexts (see Supplemental Material, Table S1). We organize retail establishments into seven NACE categories: non-specialized stores, food and beverage specialists, information and communication equipment, household equipment specialists, cultural and recreation goods, other specialized goods and food service activities (European Commission, 2008). These categories capture retail functions most embedded in daily neighbourhood life, as widely adopted in various urban retail studies (Andersson et al., 2018; Efeoğlu et al., 2024), and correspond to what Porta et al. (2012) define as secondary activities following Jacobs’ framework: local-scale commercial enterprises embedded in the daily life of neighbourhoods.

Variables: Measuring age and retail patterns

Our analysis employs four variables: mean building age and building age diversity capture age characteristics, while retail density and retail diversity measure retail patterns (Table 1). Following Jacobs’ recognition of retail patterns as a reliable indicator of broader urban vitality, we employ retail density and diversity as measurable proxies rather than direct measures of urban vibrancy.

Table 1.

Definitions and equations of the study variables.

Variable	Definition	Equation
Mean building age (MBA)	The average age of buildings within the accessible catchment	MBA _{(o; D)} = $\frac{1}{NoB} \sum_{i = 1}^{NoB} (2025 - {YEAR}_{i})$
Building age diversity (BAD)	The variety of building ages within the accessible catchment using Shannon–Weaver diversity index	BAD _{(o; D)} = −Σ (p_i × ln(p_i))
Retail density (RDen)	The total number of retail establishments within the accessible catchment	RDen _{(o; D)} = (NoPOI_i)
Retail diversity (RDiv)	The variety of retail types within the accessible catchment using Shannon-Weaver diversity index	RDiv _{(o; D)} = −Σ (p_j × ln(p_j))

o: Origin building; D: 500 m catchment; NoPOIi: number of retail points of interest; YEAR: construction year; p_i: proportion of building age category i; NoB: number of buildings within catchment; p_j: proportion of retail category j.

Source: Authors.

Each building serves as the unit of analysis for our framework. For every building, we compute all four variables by aggregating information from all buildings and retail establishments accessible within a 500-m network-based walking catchment. This accessibility-based approach ensures that age characteristics and retail patterns are measured at the same spatial resolution, reflecting each building’s experienced urban environment rather than treating buildings as isolated objects (Berghauser Pont and Marcus, 2014). This approach proves well suited for retail analysis, as people experience commercial environments through movement and proximity (Hillier et al., 1993). The 500-m distance represents the distance that most people walk for local retail activities (Gehl, 2010; Kang, 2016; Porta et al., 2010; Rao and Pafka, 2021). Buildings located near the study area boundaries inevitably have incomplete catchments, as their 500-m catchment extends beyond the available data. To address this edge effect (Gil, 2017), we removed buildings within 1 km of the study boundaries from the analysis, ensuring that every building has a complete catchment.

We employ the Shannon–Weaver diversity index for both age and retail diversity measures, calculated as:

H = - \sum (p_{i} \times \ln (p_{i}))

where p_i represents the proportion of building age categories and retail types in category i. We select this entropy-based measure because it captures both the number of categories and the evenness of their distribution. Simpler measures such as category count would not distinguish between balanced age mixing and single-period dominance, while alternatives such as the Simpson index are less responsive to rare categories. The Shannon–Weaver index weighs all categories proportionally, suiting incremental development patterns where building age periods coexist unevenly, and is widely established in urban diversity research (Rueda, 1995; Yoshimura et al., 2022).

Clustering: Identification of development types via the Temporal Fingerprint Matrix

To identify distinct temporal development patterns, we employ hierarchical clustering, which has demonstrated effectiveness in urban morphological studies for revealing underlying spatial patterns (Dibble et al., 2019; Serra et al., 2017). This approach enables data-driven classification without imposing subjective thresholds for distinguishing ‘old’ from ‘new’ or ‘diverse’ from ‘homogeneous’ areas. We standardize mean building age (MBA) and building age diversity (BAD) variables using min-max scaling to ensure equal weighting in the clustering algorithm. Given that case studies vary substantially in building numbers and age distributions, we implement balanced sampling to prevent any single city from dominating model training. We randomly sample 10,000 buildings from each city, creating a representative 30,000-building dataset, ensuring equal influence from all three cities. After evaluating multiple hierarchical clustering techniques (Kaufman and Rousseeuw, 2005), we select the complete linkage method with Euclidean distance measures based on better clustering performance and provide extensive justification for selecting the optimal cluster number in the Supplemental Material (Figure S1).

The trained clustering model is then applied to the complete dataset, assigning each building to a distinct temporal development type. We visualize this classification through our Temporal Fingerprint Matrix, which maps buildings across the bivariate space of mean building age and age diversity.

Statistical analysis: Age characteristics and retail patterns relationship

Our data structure, containing categorical development types derived from clustering alongside continuous retail density and diversity measures, necessitates the Kruskal–Wallis test for statistical analysis. This non-parametric method ranks all observations and compares rank distributions across multiple independent groups without requiring normal distribution assumptions. The Kruskal–Wallis test generates a χ² statistic that indicates whether at least one development type differs significantly from others in retail density and diversity, with the p-value determining statistical significance.

We conduct two separate tests for accessible retail density and accessible retail diversity as dependent variables, using development types as the categorical independent variable. The null hypothesis states that retail density and diversity remain consistent across all development types, suggesting no association between temporal characteristics and retail patterns. Rejecting this null hypothesis (p < 0.05) would indicate that different temporal development patterns generate significantly different retail outcomes, thereby supporting both our central research premise and Jacobs’ third condition that building age characteristics fundamentally influence retail patterns and broader urban vitality. Following the Kruskal–Wallis test, post hoc Dunn tests with Bonferroni correction were applied to identify specific pairwise differences between development types, enabling detailed comparison of retail density and diversity across all cluster combinations (see Supplemental Material, Tables S3 and S4).

Results

Description of age and retail characteristics

The spatial distribution of building age categories reveals distinct temporal signatures that validate our case study selection (Figure 3). Amsterdam exhibits a characteristic concentric pattern with historic fabric concentrated in the core, reflecting systematic expansion plans developed across centuries. Rotterdam presents minimal pre-war buildings due to wartime destruction and extensive post-war reconstruction dominating the urban fabric. Den Haag combines well-preserved historic districts with intensive recent construction concentrated in peripheral areas, particularly evident in developments after the 1990s.

Figure 3.

Spatial and statistical distribution of age categories.

These distinctive patterns are quantitatively confirmed through our independent variables (Table 2 and Figure 4). Amsterdam generates the highest mean building age (0.307) but the lowest age diversity (0.722), suggesting that preservation efforts maintained temporally homogeneous districts. Rotterdam and Den Haag show the inverse – lower mean building age but notably higher age diversity (0.855 and 0.846 respectively) – reflecting complex temporal layering created by post-war renewal and contemporary construction. Retail patterns align with these temporal characteristics: Amsterdam demonstrates the highest retail density (0.072) and diversity, Rotterdam exhibits the lowest values across both measures and Den Haag occupies an intermediate position (Table 2).

Table 2.

Descriptive statistics for building age and retail variables across case studies.

	Amsterdam (N = 163,844)				Rotterdam (N = 165,828)				Den Haag (N = 238,996)
	Min	Max	Mean	SD	Min	Max	Mean	SD	Min	Max	Mean	SD
Mean building age (scaled 0–1)	0.006	1.0	0.307	0.165	0.004	0.862	0.243	0.098	0.0	0.847	0.233	0.124
Building age diversity (Shannon)	0.0	2.0	0.722	0.474	0.0	2.02	0.855	0.439	0.0	1.96	0.846	0.474
Retail density (scaled 0–1)	0.0	0.985	0.072	0.13	0.0	0.654	0.021	0.041	0.0	1.0	0.04	0.078
Retail diversity (Shannon)	0.0	1.86	0.98	0.596	0.0	1.95	0.782	0.648	0.0	1.89	0.88	0.689

Source: Authors.

Figure 4.

Spatial and statistical distribution of the variables used in the analysis.

Identification of development types via hierarchical clustering and the Temporal Fingerprint Matrix

We applied hierarchical clustering to mean building age and building age diversity variables to identify distinct temporal development patterns and visualized results through our Temporal Fingerprint Matrix (Figure 5). To establish empirically grounded boundaries between these quadrants, we calculated the median values of building age (0.26) and building age diversity (0.41) across the entire dataset including three cities. Buildings above or below these median values are categorized as ‘high’ or ‘low’ on each axis, ensuring that quadrant boundaries reflect the observed data distribution rather than arbitrary thresholds (see Supplemental Material, Figure S2 for individual city scatter plots).

Figure 5.

Hierarchical clustering results projected on the Temporal Fingerprint Matrix and the spatial distribution of identified development types.

Hierarchical clustering identified seven distinct urban development types, revealing complexity and nuance in urban development processes that extend beyond our initial four-quadrant framework (Figure 5). Amsterdam exhibits the highest representation of incremental development with Type 1 (4.4%) and Type 7 (4.2%), concentrated primarily in the historic city core. Rotterdam presents a clear contrast, showing minimal Type 7 presence (1.2%) and complete absence of Type 1 due to wartime destruction of historical building stock. Den Haag occupies an intermediate position, lacking Amsterdam’s substantial incremental development but displaying notable Type 7 presence (4.5%) in central districts. Instant development patterns (Type 5) show increasing prevalence from Amsterdam (16.9%) to Rotterdam (18.6%) to Den Haag (19.8%), highlighting the more extensive instant development character of Rotterdam and Den Haag compared to Amsterdam’s preserved urban fabric.

To understand how these seven types differ fundamentally in their temporal characteristics, we examined building age category distributions through violin plots combined with urban tissue samples (Figure 6). Each violin plot displays the distinctive temporal ‘fingerprint’ of development processes, where wider sections indicate higher concentrations of buildings from specific historical periods.

Incremental development types (Types 1 and 7) both occupy the top-right quadrant, combining older buildings with high age diversity, yet exhibit distinctly different temporal signatures. Type 1 demonstrates a distinctive bottom-heavy violin shape, with the widest concentrations in the earliest periods (1600–1870) tapering dramatically towards newer periods. This pattern reveals urban areas where pre-industrial and early industrial buildings dominate, with only selective newer additions.

Type 7 shows a fundamentally different incremental character through its distinctive barrel shape, distributing buildings more evenly across all temporal periods. This balanced distribution creates the highest age diversity among all types while maintaining substantial older building stock. The even spread across all periods indicates that Type 7 consistently incorporates new construction alongside older fabric, embodying the ideal of ongoing, gradual urban evolution.

Type 2 (recent incremental development) occupies the top-left quadrant with high age diversity and predominantly new buildings. The violin plot shows temporal distribution beginning primarily after 1900, lacking pre-industrial building stock. Type 2 exhibits consistent temporal heterogeneity across multiple post-1900 periods, demonstrating that incremental character can emerge within compressed timeframes.

Type 3 shares similar mean building age with Type 2 but exhibits comparatively lower age diversity, placing it in a borderline position between instant and recent incremental development. Type 3 achieves greater temporal mixing than the homogeneity of instant development (such as Type 5), yet it lacks the sustained temporal heterogeneity that defines fully incremental types (such as Types 1, 2 and 7).

Type 5 (instant development) exhibits the lowest age diversity among all types, with a narrow violin shape concentrated almost entirely in the post-war reconstruction period (1946–1970 and 1971–1990). Positioned in the bottom-left quadrant with the newest buildings and lowest age diversity, Type 5 represents the archetypal instant development pattern, directly embodying Jacobs’ critique of large-scale construction approaches.

Type 6 (preserved development) occupies the bottom-right quadrant, combining older buildings with low age diversity. The violin plot reveals single-period construction dominantly concentrated in the 1916–1945 era, followed by minimal subsequent additions. This temporal signature represents areas that were built during a specific historical period and have remained largely frozen in time.

Type 4 occupies a borderline position at the intersection of the incremental and preserved development. The violin plot reveals a distinctive bimodal temporal distribution, with major concentrations in two consecutive historical periods (1871–1915 and 1916–1945), followed by very minor selective additions in later decades. Type 4 exhibits greater heterogeneity than preserved development (Type 6) through construction across two major periods yet lacks the continuous, ongoing development that defines fully incremental areas (Types 1 and 7).

Figure 6.

The comparison of identified development types via temporal fingerprints and sampled urban tissues.

The relationship between development types and retail patterns

Kruskal–Wallis tests reveal statistically significant differences across development types for both accessible retail density (χ² = 203,441, df = 6, p < 0.001) and accessible retail diversity (χ² = 102,937, df = 6, p < 0.001). Mean rank analysis clearly rejects the null hypothesis and provides robust empirical support for our central hypothesis that incremental development outperforms instant development in retail outcomes (Figure 7).

Figure 7.

Mean rank analysis results of the Kruskal–Wallis test.

The two incremental development types (Types 1 and 7) demonstrate consistently higher retail density and diversity, substantially outperforming instant development (Type 5) which ranks lowest in both retail density and diversity. Type 7 exhibits exceptional balance, ranking second in both retail density and diversity, representing the optimal combination advocated by Jacobs (1961). Type 1 achieves the highest retail density ranking among all development types. This consistent pattern confirms that incremental development creates more favourable conditions for retail vitality than instant development.

Beyond confirming the superiority of incremental over instant development types, several notable discrepancies warrant further examination. Type 1 achieves the highest retail density but ranks fourth in retail diversity. Conversely, Type 4 (hybrid development at the borderline of incremental and preserved) demonstrates the opposite pattern: ranking first in retail diversity while placing lower in density. Types 2 and 3 (recent incremental development) occupy middle positions in both measures. Interestingly, Type 6 (preserved development) outperforms not only instant development (Type 5) but also recent incremental types (Types 2 and 3) despite its low temporal heterogeneity. These patterns and their theoretical implications are examined in the discussion section.

Discussion

This study confirms our central hypothesis that development type – incremental versus instant – influences retail patterns. Our Temporal Fingerprint Matrix operationalizes building age and age diversity as integrated indicators and effectively identifies distinct urban development types. Testing Jacobs’ third condition empirically, we validate her critique that instant development across large swatches of land fails to support retail density, diversity and broader urban vitality (Jacobs, 1961). Incremental development with age-diverse buildings consistently demonstrates higher retail density and diversity, providing empirical support for this widely neglected aspect of Jacobs’ urban diversity framework. Crucially, these incrementally developed areas generate the critical mass of agglomeration that recent longitudinal studies confirm is essential for the long-term survival of street-level businesses (Kickert and Vom Hofe, 2018).

While retail density and diversity do not capture the full complexity of urban vitality, Jacobs argued that they reliably indicate the presence of broader urban diversity and vitality (Jacobs, 1961). Our findings should be interpreted within this framework: the retail patterns we identify reflect the commercial conditions generated by different temporal development processes, serving as proxies rather than comprehensive measures of urban vitality.

Four additional findings illuminate the complex relationship between age characteristics and retail patterns. First, Type 1, found exclusively in Amsterdam’s historic core, demonstrates a striking paradox: highest retail density but considerably lower retail diversity. This stems from pre-industrial building dominance (1600–1870) with minimal newer additions due to preservation policies. Detailed retail analysis (see Supplemental Material, Figure S4) confirms that Amsterdam’s historic core specializes in tourism-orientated services (restaurants, cafes, food establishments) and pushes other forms of business to the urban fringes (Koren and Hracs, 2024) rather than supporting diverse commercial ecosystems. This finding illustrates the limits of a purely economic reading of Jacobs’ third condition: in tourism-pressured historic centres, older buildings do not necessarily provide affordable space but instead attract high-rent, specialized establishments driven by agglomeration economics. Older buildings do not automatically generate retail diversity. This reinforces our argument that building age alone is insufficient and must be understood alongside age diversity to capture the full picture of how temporal development processes shape retail patterns.

Second, Type 4 (hybrid development) achieves the highest retail diversity despite poor retail density. Type 4 occupies a borderline position between incremental and preserved development, with a bimodal temporal distribution spanning two consecutive historical periods (1871–1915 and 1916–1945) plus selective later additions. Urban tissue analysis reveals that Type 4’s recent additions incorporate larger footprints potentially accommodating different retail types, such as chain stores, supermarkets and larger restaurants (Figure 6). Even under modest temporal heterogeneity, thoughtful additions to historical neighbourhoods may enhance retail diversity, supporting development strategies that emphasize ‘compatible new construction’ rather than pure preservation or comprehensive redevelopment (Preservation Green Lab, 2014). Gradual infill construction that diversifies age character, rather than instantly replacing historical fabric, enables new retail types to complement existing landscapes.

Third, recent incremental developments (Types 2 and 3) differ fundamentally from traditional incremental areas. Despite good age diversity, both rank among the lower performers in retail density and diversity, performing only better than Type 5 (instant development). This relatively poor retail density stems from the lack of older buildings, as their temporal distributions exclude the oldest building stock. These recent incremental types demonstrate that incremental character and age diversity gained over short timeframes prove insufficient for retail density and diversity. The presence of substantially older buildings remains essential for dense and diverse retail environments.

Fourth, Type 6 (preserved development) outperforms Types 2 and 3 despite its low temporal heterogeneity, demonstrating the independent value of older building stock. Even without ideal age diversity, aged buildings themselves support better retail activity than young buildings with good age diversity. This underscores that both dimensions of Jacobs’ third condition matter but the presence of substantially older buildings appears foundational. As Jacobs (1961: 188) observed, ‘the economic value of old buildings is irreplaceable at will. It is created by time’. Age diversity enhances retail density but cannot substitute for the affordability that only time can create in older buildings.

Previous studies examining building age and age diversity as separate variables have produced inconsistent findings, with some reporting positive associations with urban vitality and others finding no significant effects (De Nadai et al., 2016; Huang et al., 2023; Sung et al., 2015). Our combined approach reveals synergistic effects: age diversity creates economic stratification, while older buildings provide fundamental affordable space, enabling diverse enterprises. The Temporal Fingerprint Matrix represents a subtle and effective framework operationalizing building age and age diversity synergistically rather than separately. While our four-quadrant framework provided conceptual clarity, identifying seven distinct types reveals transitional conditions and complexity beyond simple categorizations.

The policy implications of our study underscore the fundamental value of incrementality embedded within Jacobs’ third condition. Unlike other diversity conditions achievable through direct planning interventions, building age diversity emerges through time-dependent processes that ‘cannot be rushed or preplanned’ (Grant, 2017: 98). Planning practice must therefore embrace scientifically well-informed, time-conscious and adaptive urban development strategies (Bettencourt, 2021; Romice et al., 2022). Our matrix enables planners to identify temporal character and develop tailored strategies for each context.

Incrementally developed areas require dual preservation-development strategies. Cities should preserve older buildings as time-tested solutions by reducing regulatory barriers and incentivizing adaptive reuse, while enabling compatible infill development. Economic tools like tax credits and streamlined permitting can support this balance and allow adaptive repurposing of the retail landscape in cities (Jackson et al., 2024; Powe et al., 2016). Instant development areas present the most significant challenge, since older buildings require time and cannot be created through planning. These environments demand patient transformation, preventing further homogenization through varied building types and ‘efficient phasing’, where development occurs sequentially through multiple agents over time (Adams et al., 2013; Love and Crawford, 2011). This gradual approach is vital, as recent research demonstrates that sudden physical or commercial overhauls can destabilize collective efficacy by outpacing social adaptation (Zahnow et al., 2022).

Recent incremental areas lacking older buildings offer more promise for gradual improvement than instant development areas. As Jacobs (1961) observed, time transforms today’s new buildings into tomorrow’s affordable structures. Cities should prevent homogenization in these environments while supporting adaptive reuse and maintaining incremental character. Preserved areas require strategic infill, introducing compatible construction without compromising historic fabric. We should note that pure preservation creates affordability traps, as preservation policies often increase property values rather than maintaining accessible spaces, as Jacobs envisioned (Been et al., 2014; Glaeser, 2012). Jacobs advocated plain, ordinary, low-value old buildings complemented by different ages, not expensive preserved districts. In this sense, strategic diversification maintains economic accessibility, which is essential for dense and diverse retail patterns.

Several limitations suggest future research directions. Our findings raise questions about optimal ratios and tipping points where age diversity benefits diminish. Understanding what proportion of older versus newer buildings maximizes retail density and diversity warrants further investigation across different contexts. Several methodological refinements could enhance future analyses: multi-scalar accessibility analysis beyond our 500-m threshold; alternative clustering algorithms; building area weighting in age calculations, as building sizes significantly influence measures; and three-dimensional analysis incorporating height, as taller buildings accommodate more units and activities per footprint.

Moreover, our analysis measures retail presence through establishment locations and categorical diversity rather than dynamic indicators such as footfall data, sales volume or real-time activity patterns. While this POI-based approach is common practice in studies examining Jacobs’ framework (Gómez-Varo et al., 2022), future research could integrate dynamic activity measures such as walking activity (Sung et al., 2015), social media activity (Huang et al., 2023) or sales volumes (Yoshimura et al., 2022) to examine whether the temporal development patterns identified here also influence real-time urban activity. Finally, research should expand beyond retail to examine social, cultural and innovation indicators, refining understanding of how temporal development processes shape broader urban vitality and guide future development and preservation policies.

Conclusion

This study examined how building age characteristics affect retail density and diversity patterns in Amsterdam, Rotterdam and Den Haag. To that end, we developed the Temporal Fingerprint Matrix, which combines mean building age and building age diversity to distinguish between incremental and instant development patterns. Our analysis demonstrates that incrementally developed areas – characterized by high age diversity and substantial older building stock – consistently outperform instantly developed areas with uniform building ages. These findings validate Jacobs’ argument that large-scale instant development undermines the conditions necessary for vibrant commercial districts, whereas incremental development with mixed building ages creates the essential foundation for diverse and concentrated retail activities.

Our research contributes to three critical knowledge gaps in urban studies. First, we provide the first systematic empirical test of how building age characteristics influence retail patterns, filling the long-standing research gap surrounding Jacobs’ most neglected diversity condition. Second, we introduce an integrated analytical framework that treats building age and diversity as complementary indicators, resolving the previous studies’ methodological limitation of examining these variables in isolation. Third, we deliver evidence-based insights demonstrating how distinct development types (from incremental to instant) shape retail patterns and derive actionable policy recommendations for urban planning practice. These contributions illustrate how rigorous empirical testing of foundational urban theories like Jacobs’ diversity framework can simultaneously advance theoretical knowledge and practical applications in urban studies.

Supplemental Material

sj-docx-1-usj-10.1177_00420980261462840 – Supplemental material for Building age diversity and retail patterns: A morphometric exploration of Jacobs’ neglected condition through the Temporal Fingerprint Matrix

Supplemental material, sj-docx-1-usj-10.1177_00420980261462840 for Building age diversity and retail patterns: A morphometric exploration of Jacobs’ neglected condition through the Temporal Fingerprint Matrix by Onur Tümtürk, Hulusi Eren Efeoğlu, Mert Akay and Olgu Çalışkan in Urban Studies

Footnotes

ORCID iDs

Onur Tümtürk

Hulusi Eren Efeoğlu

Mert Akay

Olgu Çalışkan

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Onur Tümtürk acknowledges personal funding from the Fulbright Postdoctoral Scholarship Program (FY-2025-TR-PD-04), held at the Massachusetts Institute of Technology, Senseable City Lab, during the writing of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

The building age data used in this study are publicly available from the Netherlands’ BAG (Basisregistratie Adressen en Gebouwen) public registry (https://www.kadaster.nl/about-us), and retail data were obtained from OpenStreetMap’s POI dataset (). Analysis code is available from the corresponding author upon reasonable request.

Supplemental material

Supplemental material for this article is available online.

References

Adams

Tiesdell

White

(2013) Smart parcelization and place diversity: Reconciling real estate and urban design priorities. Journal of Urban Design 18(4): 459–477.

Andersson

Larsson

Wernberg

(2018) Urban preferences, amenities and age: Exploring the spatial distribution of age in Stockholm from 1991 to 2011. Regional Science Policy & Practice 10(4): 367–382.

Bantman-Masum

(2020) Unpacking commercial gentrification in central Paris. Urban Studies 57(15): 3135–3150.

Been

Ellen

McCabe

(2014) Preserving History or Hindering Growth? The Heterogeneous Effects of Historic Districts on Local Housing Markets in New York City. Working paper. National Bureau of Economic Research.

Berghauser Pont

Marcus

(2014) Innovations in measuring density: From area and location density to accessible and perceived density. Nordic Journal of Architectural Research 26(2): 11–30.

Bettencourt

LMA

(2021) New empirically driven approaches to inclusive human development in cities. In: Kumar

Meshram

(eds) Sustainable Development Goals and Indian Cities, 1st edn. Routledge India, pp.119–136.

De Nadai

Staiano

Larcher

, et al. (2016) The death and life of great Italian cities: A mobile phone data perspective. In: Proceedings of the 25th international conference on world wide web, Montréal, QC, Canada, 11 April 2016, pp.413–423.

Delclòs-Alió

Miralles-Guasch

(2018) Looking at Barcelona through Jane Jacobs’s eyes: Mapping the basic conditions for urban vitality in a Mediterranean conurbation. Land Use Policy 75: 505–517.

Dibble

Prelorendjos

Romice

, et al. (2019) On the origin of spaces: Morphometric foundations of urban form evolution. Environment and Planning B: Urban Analytics and City Science 46(4): 707–730.

10.

Efeoğlu

Joutsiniemi

Mozuriunaite

(2024) Exploring the plot patterns of the retail landscape: The case of the Helsinki Metropolitan area. Environment and Planning B: Urban Analytics and City Science 51(6): 1210–1226.

11.

European Commission (2008) NACE Rev. 2: Statistical Classification of Economic Activities in the European Community. Publications Office.

12.

Ewing

Cervero

(2010) Travel and the built environment: A meta-analysis. Journal of the American Planning Association 76(3): 265–294.

13.

Gehl

(1987) Life Between Buildings: Using Public Space. Van Nostrand Reinhold.

14.

Gehl

(2010) Cities for People. Island Press.

15.

Gil

(2017) Street network analysis “edge effects”: Examining the sensitivity of centrality measures to boundary conditions. Environment and Planning B: Urban Analytics and City Science 44(5): 819–836.

16.

Glaeser

(2012) Triumph of the City: How Our Greatest Invention Makes Us Richer, Smarter, Greener, Healthier, and Happier. Penguin Books.

17.

Gómez-Varo

Delclòs-Alió

Miralles-Guasch

(2022) Jane Jacobs reloaded: A contemporary operationalization of urban vitality in a district in Barcelona. Cities 123: 103565.

18.

Grant

(2017) Time, scale, and control: How new urbanism mis(uses) Jane Jacobs. In: Page

Mennel

(eds) Reconsidering Jane Jacobs, 1st edn. Routledge, pp.91–176.

19.

Hillier

Penn

Hanson

, et al. (1993) Natural movement: Or, configuration and attraction in urban pedestrian movement. Environment and Planning B: Planning and Design 20(1): 29–66.

20.

Hirt

Zahm

(2012) The Urban Wisdom of Jane Jacobs. Routledge.

21.

Huang

Cui

, et al. (2023) Re-examining Jane Jacobs’ doctrine using new urban data in Hong Kong. Environment and Planning B: Urban Analytics and City Science 50(1): 76–93.

22.

Jackson

Lawson

Orr

, et al. (2024) Repurposing retail space: Exploring stakeholder relationships. Urban Studies 61(1): 148–164.

23.

Jacobs

(1961) The Death and Life of Great American Cities. Random House.

24.

Kang

C-D

(2016) Spatial access to pedestrians and retail sales in Seoul, Korea. Habitat International 57: 110–120.

25.

Kaufman

Rousseeuw

(2005) Finding Groups in Data: An Introduction to Cluster Analysis. Wiley Series in Probability and Statistics, 1st edn. Wiley.

26.

Kickert

Vom Hofe

(2018) Critical mass matters: The long-term benefits of retail agglomeration for establishment survival in downtown Detroit and The Hague. Urban Studies 55(5): 1033–1055.

27.

King

(2013) Jane Jacobs and ‘the need for aged buildings’: Neighbourhood historical development pace and community social relations. Urban Studies 50(12): 2407–2424.

28.

Klemek

(2007) Placing Jane Jacobs within the transatlantic urban conversation. Journal of the American Planning Association 73(1): 49–67.

29.

Koren

Hracs

(2024) Negotiating the night: How nightclub promoters attune their curatorial practices to the intra-urban dispersal of nightlife in Amsterdam. Urban Studies 61(12): 2328–2348.

30.

Long

Huang

(2019) Does block size matter? The impact of urban design on economic vitality for Chinese cities. Environment and Planning B: Urban Analytics and City Science 46(3): 406–422.

31.

Love

Crawford

(2011) Plot logic: Character-building through creative parcelisation. In: Tiesdell

Adams

(eds) Urban Design in the Real Estate Development Process, 1st edn. Wiley, pp.92–113.

32.

Mccarthy

(1999) The redevelopment of Rotterdam since 1945. Planning Perspectives 14(3): 291–309.

33.

Meltzer

Capperis

(2017) Neighbourhood differences in retail turnover: Evidence from New York City. Urban Studies 54(13): 3022–3057.

34.

Montgomery

(1998) Making a city: Urbanity, vitality and urban design. Journal of Urban Design 3(1): 93–116.

35.

Nabielek

Kronberger-Nabielek

Hamers

(2013) The rural-urban fringe in the Netherlands: Recent developments and future challenges. SPOOL 1(1): 1–18. https://doi.org/10.7480/spool.2013.1.624.

36.

Porta

Latora

Strano

(2010) Networks in urban design. Six years of research in multiple centrality assessment. In: Estrada

Fox

Higham

, et al. (eds) Network Science. Springer London, pp.107–129.

37.

Porta

Latora

Wang

, et al. (2012) Street centrality and the location of economic activities in Barcelona. Urban Studies 49(7): 1471–1488.

38.

Powe

Mabry

Talen

, et al. (2016) Jane Jacobs and the value of older, smaller buildings. Journal of the American Planning Association 82(2): 167–180.

39.

Preservation Green Lab (2014) Older, Smaller, Better Measuring How the Character of Buildings and Blocks Influences Urban Vitality. National Trust for Historic Preservation.

40.

Rao

Pafka

(2021) Shopping morphologies of urban transit station areas: A comparative study of central city station catchments in Toronto, San Francisco, and Melbourne. Journal of Transport Geography 96: 103156.

41.

Romice

Rudlin

AlWaer

, et al. (2022) Urban design as a specialised, evidence-based, coordinated educational and professional endeavour. Proceedings of the Institution of Civil Engineers – Urban Design and Planning 175(4): 179–198.

42.

Rueda

(1995) Ecologia urbana: Barcelona i la seva regió metropolitana com a referents. Beta Editorial.

43.

Rutte

Abrahamse

(2016) Atlas of the Dutch Urban Landscape: A Millenium of Spatial Development. Thoth.

44.

Savini

Boterman

Van Gent

WPC

, et al. (2016) Amsterdam in the 21st century: Geography, housing, spatial development and politics. Cities 52: 103–113.

45.

Serra

Gil

Pinho

(2017) Towards an understanding of morphogenesis in metropolitan street-networks. Environment and Planning B: Urban Analytics and City Science 44(2): 272–293.

46.

Southworth

Ben-Joseph

(1997) Streets and the Shaping of Towns and Cities. Island Press.

47.

Sung

Lee

Cheon

(2015) Operationalizing Jane Jacobs’s urban design theory: Empirical verification from the great city of Seoul, Korea. Journal of Planning Education and Research 35(2): 117–130.

48.

Talen

(2010) The context of diversity: A study of six Chicago neighbourhoods. Urban Studies 47(3): 486–513.

49.

Van der Valk

(2015) Town planning in the Netherlands since 1800: Responses to enlightenment ideas and geopolitical realities. International Planning Studies 20(3): 316–317.

50.

Wagenaar

(2016) Town Planning in the Netherlands: Since 1800. Nai010 Publishers.

51.

White

Hickie

Orr

, et al. (2023) The experience economy in UK city centres: A multidimensional and interconnected response to the ‘death of the high street’? Urban Studies 60(10): 1833–1852.

52.

Ren

, et al. (2018) Check-in behaviour and spatio-temporal vibrancy: An exploratory analysis in Shenzhen, China. Cities 77: 104–116.

53.

Liu

(2018) How block density and typology affect urban vitality: An exploratory analysis in Shenzhen, China. Urban Geography 39(4): 631–652.

54.

Yoshimura

Kumakoshi

Milardo

, et al. (2022) Revisiting Jane Jacobs: Quantifying urban diversity. Environment and Planning B: Urban Analytics and City Science 49(4): 1228–1244.

55.

Yue

Zhuang

Yeh

AGO

, et al. (2017) Measurements of POI-based mixed use and their relationships with neighbourhood vibrancy. International Journal of Geographical Information Science 31(4): 658–675.

56.

Zahnow

Corcoran

Kimpton

, et al. (2022) Neighbourhood places, collective efficacy and crime: A longitudinal perspective. Urban Studies 59(4): 789–809.

57.

Zeng

Song

, et al. (2018) Spatially explicit assessment on urban vitality: Case studies in Chicago and Wuhan. Sustainable Cities and Society 40: 296–306.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

4.72 MB