Integrated Management of Urban Landscape and Public Security

Abstract

Criminality represents one of the major challenges faced by Brazilian cities, particularly due to the limited state capacity to control violence. To provide support for addressing this issue, the research focuses on systematizing responses to the investigative question regarding the identifiable foundations for sharing urbanistic-criminal solutions within the framework of the urban agenda. With the aim of analyzing this reality through a case study in Itumbiara, Goiás, Brazil, the assessment relied on field surveys and relational analysis of criminal, landscape, and institutional data. Considering three temporal intervals—before, parallel, and after the coronavirus disease 2019 (COVID-19) pandemic—to recognize its influence on the phenomenon under review, the results on the distribution of thefts and robberies along a road sample show a decline in crime even after the official end of the health crisis. Socio-behavioral and socio-political parameters scored lower than socio-morphological ones but overall were classified as medium-low quality. The formulated guidelines combine analytical variables and criminological theories to support public policies for integrated management of safe scenarios, reinforcing the hypothesis that such interactions enable principles for coordinated measures in planning and security in urbanized areas.

1. Introduction

Directly linked to the state’s capacity for controlling violence, public security stands as one of the main indicators of social performance. The absence or ineffectiveness of policies aimed at its reduction intensifies instability and chaos within society [1,2]. The mitigation of urban criminality, including through spatial solutions, constitutes a critical global challenge [3]. In pursuit of this goal, several criminological theories, many of them linked to environmental planning, have been developed.

Aiming at the formulation of distinctive and original contributions of this research, the theoretical inputs of interest are especially connected to socio-behavioral issues, more directly related to offenders’ conduct, commonly influenced by societal practices; socio-morphological aspects, referring to features of urbanized landscapes that either facilitate or restrain criminal acts; and socio-political dimensions, aligned with processes of city governance and civic engagement. Evidently, such categorization is not definitive, since criminality constitutes a complex phenomenon characterized by multidimensional interactions [4]. It is, therefore, a generic systematization of the classical literature and contemporary references of the principal focal points within the main theories relevant to the subject under examination.

In the case of socio-behavioral issues, the Routine Activity theoretical framework emphasizes that illicit incidents emerge from the simultaneous occurrence of the existence of a vulnerable target, the absence of protective oversight, and the presence of a motivated perpetrator [5,6,7,8]. Complementarily, the Criminal Opportunities perspective underscores that individuals display limited inclination to maintain preventive dispositions toward illicit practices [9,10]. The Marginal Labeling lens further elucidates the mechanisms through which subjects are redefined according to externally imposed identities [11,12,13]. Building upon these insights, the Criminal Economy approach interprets deviant conduct through reasoned decision schemata, wherein actors weigh prospective benefits against anticipated drawbacks before acting [11,14]. This reasoning parallels Rational Choice postulates, which maintain that engagement in unlawful behavior results from comparative evaluations of expected advantages relative to possible sanctions [5,6,11,15,16,17,18]. Finally, Social Disorganization bases highlight the heightened probability of illegal expression in communities destabilized by weakened traditions and restricted access to collective resources [19,20].

Within socio-morphological aspects, the Locational Criminality theoretical model contends that deviant practices are linked to opportunities structured and facilitated by spatial configurations [21,22,23]. In turn, Personal Space analysis demonstrates that immediate surroundings exert influence on predispositions toward illicit behavior [24,25,26]. Social Cohesion study posits the premise that environments fostering urban continuity reinforce solidarity while simultaneously creating conditions conducive to criminal expression [27,28]. Territorial Defense perspectives emphasize the relevance of spatial context for collective maintenance and shared responsibility [29,30].

The Social Density standpoint proposes that elevated constructive population rates stimulate interactive tendencies, accompanied by regulatory mechanisms of both environmental and behavioral nature [31,32]. Complementarily, the Safe Spaces approach asserts that settings characterized by fluid access and enhanced mobility encourage the emergence of ‘natural guardians’ [9,24,25,33]. Among these formulations, the Environmental Design paradigm stands out as a foundational construct, underpinning the Crime Prevention through Environmental Design (CPTED) framework [15,34]. Its principles align with the Universal Design concept, which underscores the imperative of accessibility across spatial configurations [35,36].

The Defensible Space construct advocates for arrangements incorporating protective functions explicitly implemented by community members [34,37,38,39]. Concurrently, the Spatial Needs lens maintains that built environments shape perceptions of safety [24,25,40], while the Spatial Victimization frame reveals that locations of prolonged social interaction facilitate the identification of potential targets [24,25,41].

Regarding socio-political dimensions, Social Defense theoretical premises establish that public safety is conceived as a responsibility extending beyond governmental authority [42,43]. On the other hand, Social Control propositions argue that collective accountability for societal equilibrium is exercised through the preservation of explicit norms, which function as fundamental mechanisms of deterrence [44,45,46,47,48]. The Natural Surveillance framework, widely acknowledged, situates individuals as ‘guardians’ of communal spaces [9,49]. Broken Windows tenets contend that manifestations of physical disorder contribute to the intensification of unlawful practices [40,50,51,52], while Results-Based Management models emphasize the necessity of systematic evaluation of security performance, covering both strategic initiatives and tactical measures [53,54], supported by Criminal Mapping procedures for data visualization on crimes to enhance security governance [24,25,55].

Concertedly, the articulation of these theoretical formulations converges to demonstrate the interrelation of behavioral, morphological, and political conditions in sustaining social and urbanistic stabilities while mitigating criminogenic dynamics. In association with empirical outcomes derived from a singular methodological experiment developed, the study yields distinct contributions and innovations relative to the international background.

In the Brazilian perspective, criminality is constitutive of the country’s urban context [56,57]. This assertion is reinforced by institutional data, which further highlights that the issue is more significant than official records suggest due to under-reporting [58]. Because of its morpho-functional characteristics, the urbanized space can both deter and facilitate criminal acts [15,16,17,34,37,38]. Consequently, its planning and upkeep play crucial roles in spatial security and local defense conducted by citizens themselves [9,34,37,38,39].

Despite progress in sectoral cooperation, significant gaps remain in the integrated management of Brazilian cities [58]. This scenario encompasses public safety and urban landscape concerns, particularly in open spaces, where theft and robbery are common crimes—respectively involving the unlawful taking of property without and with serious threat or violence against the victim [59].

Specifically in the case of Itumbiara, Goiás, Brazil, during the two years preceding the coronavirus disease (COVID-19) pandemic—namely 2018 and 2019—these offenses amounted to 3520 entries (73.5% thefts and 26.5% robberies). In the subsequent biennium marked by the health emergency, 2020 and 2021, the tally fell to 2574, reflecting a 26.9% decline, with thefts representing 80.6% and robberies 19.4% of the aggregate. Paradoxically, in the two years following the crisis (2022 and 2023), the totals continued to diminish, reaching 2140 occurrences (86.2% thefts and 13.8% robberies). These data indicate reductions of 39.2% and 16.9% compared to the first and second intervals, respectively [60].

The spatial distribution of these incidents across the city’s landscape varied according to the type of crime. Thus, in every period, thefts tended to accumulate in central districts, whereas robberies were more scattered, despite their higher concentration in the downtown area. In seeking to understand this issue, the research aims to analyze the study objects—defined as public urban spaces in Itumbiara, particularly streets—as a basis for policies aimed at integrated management of a safe urban environment. Its implementation is enabled by addressing the following investigative question: what identifiable underpinnings exist in these essential locations that support a collective process of sharing urbanistic-criminal solutions within the framework of the urban agenda?

The set of investigative findings is therefore directed toward testing the hypothesis that interactions between criminal and landscape variables enable principles for coordinated measures in planning and security in urbanized areas for the integrated management of public safety and the urban environment. To this end, a dedicated methodological essay is structured to guide the development of research.

2. Materials and Methods

With a quali-quantitative approach applied, the study was guided by exploratory, descriptive, and synthetic–analytical methods, supported by field survey techniques and relational analysis of criminal, landscape, and institutional data. As previously mentioned, the research area encompasses the city of Itumbiara, specifically its districts, and the objects under investigation correspond to public spaces, with an emphasis on streets.

2.1. Research Area

Located in Brazil’s Central-West Region, the municipal territory spans 2447 km² and is projected to have 113,322 inhabitants by 2025, with nearly 97% living in urbanized areas [61]. It lies in the south-central portion of the state of Goiás (Figure 1). The urban territory of the municipal seat, with an approximate extent of 70.3 km², is in the east-central portion of the municipality [62]. It is composed of 83 districts, which more closely correspond to plots resulting from land subdivisions. Following these delimitations, the spatial units for investigation were defined.

2.2. Study Units

For the selection of specific analytical objects (street segments), the district with the highest incidence of thefts and robberies was preliminarily identified (Setor Central), considering the before (2018–2019), parallel (2020–2021), and after (2022–2023) periods of the COVID-19 pandemic. These data were obtained for restricted use from the Public Security Observatory of the State of Goiás (Observatório de Segurança Pública do Estado de Goiás—OSP-GO) [60], and during their tabulation in Excel for Windows 365 (Microsoft Corp., Redmond, WA, USA), values corresponding to rural zones (outside the official urban perimeter), as well as other inconsistencies, were excluded.

Using the statistical software R 4.6.0 (R Foundation for Statistical Computing, Vienna, Austria) to test whether the reduction in thefts and robberies across the intervals before, parallel, and after the pandemic is statistically significant, the Friedman Test was applied as the nonparametric alternative to repeated-measures of Analysis of Variance (ANOVA), suitable for comparing three or more time points measured on the same experimental units—in this case, the same districts observed across the three periods—without requiring data normality or homogeneity of variances. These procedures are followed by pairwise Wilcoxon Signed-Rank comparisons with Holm Correction—sequential p-value adjustment [65,66].

These assumptions are unlikely given the small number of selected districts (n = 12) and the presence of a clear outlier (Setor Central), whose volume of thefts and robberies is several times higher than that of other neighborhoods. The null hypothesis of the test is that the distribution of case counts is the same across the three periods; a p-value below 0.05 indicates that at least one period differs statistically from the others, without specifying which. When the global test is significant, pairwise comparisons are conducted using the Wilcoxon Signed-Rank test, appropriate because each district is measured in all three periods, generating dependent (paired) observations. Since three comparisons are made for each type of crime (before versus pandemic—2018–2019 versus 2020–2021; pandemic versus after—2020–2021 versus 2022–2023; and before versus after—2018–2019 versus 2022–2023), p-values are adjusted using the Holm method, which controls the family-wise Type I error rate while maintaining statistical validity. Effect size, which measures the magnitude of distributional differences across time, is reported by Kendall’s W for the global test (ranges from 0 to 1; values above 0.5 are considered large effects) and by Wilcoxon’s r for each pairwise comparison (|r| < 0.3: small effect; 0.3–0.5: moderate effect; >0.5: large effect) [65,66].

Subsequently, the mapping of streets was conducted to spatialize criminal points across the three temporal intervals (2018–2019, 2020–2021, and 2022–2023). These procedures were conducted using geoprocessing software such as Arc Geographic Information System (ArcGIS) Pro 3.6.4 (Environmental Systems Research Institute–ESRI, Redlands, CA, USA) and Quantum Geographic Information System (QGIS) 3.4 (Open Source Geospatial Foundation, Beaverton, OR, USA), along with online platforms like Google Earth Pro 7.3 (Google LLC, Mountain View, CA, USA). This enabled the visualization and analysis of geographic data, supporting the generation of heat maps in those programs, based on crime hotspots (active points) recorded by statistics from that observatory [60]. Due to the imprecision of municipal georeferenced databases, vector files of streets and districts boundaries were created in the same geoprocessing software, through the production of shapefiles within a Geographic Information System (GIS) environment.

For generating the heat maps, the Kernel Density Estimation (KDE) was executed with an output cell dimension of 7.42791878343212 × 10⁻⁵ and the search radius applied corresponded to a default value of 0.0011039395508422034 degrees. Spatial quantification was expressed in square map units, and the resulting raster contained density values as its primary output. The computational approach adopted was the planar method, ensuring that calculations were performed within a two-dimensional Cartesian framework.

The Getis-Ord Gi\* statistic method is used as a local indicator of spatial autocorrelation. For each street, it compares the sum of observed values (the number of thefts or robberies) on that street and its districts with the value expected under a random spatial distribution of cases. The result is expressed, for each street, by a z-score and an associated p-value. Positive and statistically significant z-scores indicate a clustering of high values around that street (hot spot), whereas negative and significant z-scores indicate a clustering of low values (cold spot). The absence of statistical significance indicates that the observed value does not differ from what would be expected under spatial randomness [67]. The software classifies each street into seven categories (Gi_Bin column), ranging from −3 (cold spot with 99% confidence) to +3 (hot spot with 99% confidence), with 0 representing no statistically significant spatial pattern.

In addition to the heat maps for spatial recognition of thefts and robberies across the three defined temporal intervals, aimed at selecting the most affected street in the Setor Central (Avenida Beira-Rio), the specific data of this roadway were also interpreted quantitatively and proportionally by months of the year and by time periods of the day. Using the Python 3.14 computational environment (Python Software Foundation, Wilmington, DE, USA), the nonparametric statistical test of Spearman’s Correlation Coefficient (ρ) was also applied, which assesses monotonic relationships based on variable rankings [68]. The adoption of this approach is justified by its greater robustness for small samples and in cases where the data contains extreme values (outliers), as observed in certain specific years of records. The strength of association was interpreted according to the Cohen Intervals: 0.10–0.29 (weak), 0.30–0.49 (moderate), and 0.50–1.00 (strong) [69].

These sets of procedures for interpreting the study units aimed at deepening the understanding of criminal behavior within the local analyzed, concurrently with the assessment of specific parameters.

2.3. Analytical Variables

For the interpretation of relevant characteristics, the variables systematized in

Table 1. A synthesis panel of descriptions and justifications of the analytical variables and the respective parameters adopted. Sources adapted: [70] and other references consulted.

Variables	Parameters	Descriptions and Justifications
Socio-Behavioral
Evasion	Escape routes	Short blocks unfavorable to police pursuit (due to the proximity of street corners) and other spatial elements structuring exit and dispersal paths for offenders (common in parks, squares, and similar places), which facilitate their withdrawal from the field of vision and therefore restrict their apprehension.
Concealment	Physical barriers	Spatial components that form physical or visual hiding places for offenders, enabling surprise attacks without giving victims time to react.
Shielding	Visual obstructions
Socio-Morphological
Dimensions	Street scale	Proportions of sidewalks that optimize pedestrian circulation, configured as spatial guardians.
	Urban scale	Proportions of built elements inviting the attraction of people and their spatial perception for natural surveillance.
Densities	Constructive/population densification	Higher levels of residential building occupation generally correspondent to a greater number of users in outdoor spaces surrounding constructions.
	Diversity of uses	Wider range of land uses that distinguishes distinct groups of users and enhances urban vitality and public safety.
Accessibilities	Relations between public and private spaces	Potential for access and circulation associated with the movement of people, fostering socio-spatial interactions.
	Spatial intervisibility	Facades and enclosures with transparent openings that induce interactions and visual access, enabling spatial monitoring at the observer’s eye level.
Attractiveness	Local references	Presence of distinctive elements (nodal points and/or landmarks) that foster recognition and attract individuals and groups.
	Public lighting	Spacing of permanent nighttime light sources to ensure adequate distribution of luminosity in public spaces.
Permanencies	Arrangement of seating and/or stopping furniture	Potential for spatial appropriation by users due to furniture elements.
	Shading for seating and/or stopping	Potential for environmental comfort that encourages users to remain in the location.
Socio-Political
Institutional instruments/ tactical tools	Maintenance of public spaces	Condition of conservation of streets (especially sidewalks) and other public components, revealing physical–territorial order/disorder.
	Monitoring of public and private spaces	Existence of permanent personnel and control devices in places of common and private use.

were considered, according to their descriptive and explanatory approaches. These parameters were further subdivided into socio-behavioral, related to public security (criminal practices of evasion, concealment, and shielding); socio-morphological, concerning the urban landscape (formal and functional conditions of dimensions, densities, accessibilities, attractiveness, and permanencies); and socio-political (institutional instruments and tactical tools).

In

Table 2. A synthesis panel of the classification of analytical variables. Sources adapted: [70] and other references consulted (see Appendix A).

Variables/ Parameters			Classes (Unit of Analysis: Length of the Street Segment)			Possible Reducers	Possible Promoters
Socio-Behavioral
Evasion	Escape routes: average percentage		less than 10%	high quality	tendency for segment data located within the lower decile: potential for substantial minimization of adverse escape routes characteristics	Short blocks and other elements facilitating exit and dispersion (e.g., parks, squares, and similar spaces)	Long blocks and possible barriers to exit and dispersion
			from 10 to 25%	medium-high quality	tendency for remaining segment data in the lower decile located within the lower quartile: potential for relative minimization of adverse escape routes characteristics
			from 26 to 50%	medium-low quality	tendency for segment data located within the intermediate-low quartile: potential for relative maximization of adverse escape routes characteristics
			more than 50%	low quality	tendency for segment data located within the two upper quartiles: potential for substantial maximization of adverse escape routes characteristics
Concealment	Physical barriers: average percentage		less than 10%	high quality	tendency for segment data located within the lower decile: potential for substantial minimization of adverse physical barriers characteristics	Large dimensions of physical barriers (including shrubbery and similar elements)	Absence of physical barriers
			from 10 to 25%	medium-high quality	tendency for remaining segment data in the lower decile located within the lower quartile: potential for relative minimization of adverse physical barriers characteristics
			from 26 to 50%	medium-low quality	tendency for segment data located within the intermediate-low quartile: potential for relative maximization of adverse physical barriers characteristics
			more than 50%	low quality	tendency for segment data located within the two upper quartiles: potential for substantial maximization of adverse physical barriers characteristics
Shielding	Visual obstructions: average percentage		less than 10%	high quality	tendency for segment data located within the lower decile: potential for substantial minimization of adverse visual obstructions characteristics	Large dimensions of visual obstructions (including shrubbery and similar elements)	Absence of visual obstructions
			from 10 to 25%	medium-high quality	tendency for remaining segment data in the lower decile located within the lower quartile: potential for relative minimization of adverse visual obstructions characteristics
			from 26 to 50%	medium-low quality	tendency for segment data located within the intermediate-low quartile: potential for relative maximization of adverse visual obstructions characteristics
			more than 50%	low quality	tendency for segment data located within the two upper quartiles: potential for substantial maximization of adverse visual obstructions characteristics
Socio-Morphological
Dimensions	Street scale: average proportion (sidewalk width/roadway width) 1		more than 1/1 (more than 1.00)	high quality	framing of segment proportional average between sidewalk and roadway widths with potential for significant maximization of positive street-scale attributes	Absence of sidewalks or pavement with obstructions to pedestrian passage	Pavement with a higher proportion of sidewalks without obstructions to pedestrian passage or expansion of sidewalk into open areas
			from 1/1 to 1/2 (from 1.00 to 0.50)	medium-high quality	framing of segment proportional average between sidewalk and roadway widths with potential for relative maximization of positive street-scale attributes
			from 1/2 to 1/4 (from 0.49 to 0.25)	medium-low quality	framing of segment proportional average between sidewalk and roadway widths with potential for relative minimization of positive street-scale attributes
			less than 1/4 (less than 0.25)	low quality	framing of segment proportional average between sidewalk and roadway widths with potential for substantial minimization of positive street-scale attributes
	Urban scale: average proportion (block length/higher building height) 2		more than 1/1 (more than 1.00)	high quality	framing of segment proportional average between block length and higher building height with potential for significant maximization of positive urban scale attributes	Components of human scale mismatch	Components of human scale adjustment
			from 1/1 to 1/2 (from 1.00 to 0.50)	medium-high quality	framing of segment proportional average between block length and higher building height s with potential for relative maximization of positive urban scale attributes
			from 1/2 to 1/4 (from 0.49 to 0.25)	medium-low quality	framing of segment proportional average between block length and higher building height with potential for relative minimization of positive urban scale attributes
			less than 1/4 (less than 0.25)	low quality	framing of segment proportional average between block length and higher building height with potential for substantial minimization of positive urban scale attributes
Densities	Constructive/ population densification: average constructive population density 3		more than 36 inh./ha	high quality	framing of segment average constructive population density above twice the city mean: potential for significant expansion of natural surveillance	Absence of residential buildings or vacant lots	Uses related to large audiences
			from 24 inh./ha to 38 inh./ha	medium-high quality	framing of segment average constructive population density between city mean and twice this value: potential for relative expansion of natural surveillance
			from 12 inh./ha to 24 inh./ha	medium-low quality	framing of segment average constructive population density between city mean and half this value: potential for relative reduction in natural surveillance
			less than 12 inh./ha	low quality	framing of segment average constructive population density below half the city mean: potential for significant reduction in natural surveillance e
	Diversity of uses: degree of land use diversity		more than 3 uses	high quality	greater degree of segment land use diversity: potential for significant urban vitality	Vacant lots	Diversified uses in the same location (e.g., residential, commercial, institutional, services, leisure, and others)
			3 uses	medium-high quality	median degree of segment land use diversity: potential for relative urban vitality
			2 uses	medium-low quality	smaller degree of segment land use diversity: potential for reduced urban vitality
			1 use	low quality	absence of segment land use diversity: potential for very reduced urban vitality
Accessibilities	Relations between public and private spaces: average percentage (free spatial access)		more than 50%	high quality	tendency for segment data located within the two upper quartiles: potential for substantial maximization of positive relations between public and private spaces	Marked barriers between public and private spaces, including controlled accesses	Projections of private uses over free-access setbacks or over public spaces
			from 26 to 50%	medium-high quality	tendency for segment data located within the intermediate-upper quartile: potential for relative maximization of positive relations between public and private spaces
			from 10 to 25%	medium-low quality	tendency for remaining segment data in the lower decile located within the lower quartile: potential for relative minimization of positive relations between public and private spaces
			less than 10%	low quality	tendency for segment data located within the lower decil: potential for substantial minimization of positive relations between public and private spaces
	Spatial intervisibility average percentage (openings with transparency)		more than 50%	high quality	tendency for segment data located within the two upper quartiles: potential for substantial maximization of positive spatial intervisibility attributes	High proportion of openings above the observer’s eye level	High proportion of active facades at ground level
			from 26 to 50%	medium-high quality	tendency for segment data located within the intermediate-upper quartile: potential for relative maximization of positive spatial intervisibility attributes
			from 10 to 25%	medium-low quality	tendency for remaining segment data in the lower decile located within the lower quartile: potential for relative minimization of positive r spatial intervisibility attributes
			less than 10%	low quality	tendency for segment data located within the lower decil: potential for substantial minimization of positive r spatial intervisibility attributes
Attractiveness	Local references: degree of presence		more than 3 references	high quality	greater degree of segment local references presence: potential for significant local recognition and attractiveness	References without potential to attract large audiences	References with potential to attract large audiences
			3 references	medium-high quality	median degree of segment local references presence: potential for relative local recognition and attractiveness
			2 references	medium-low quality	lower-median degree of segment local references presence: potential for reduced local recognition and attractiveness
			1 or absence of references	low quality	smaller degree of segment local references presence or their absence: potential for very reduced local recognition and attractiveness
	Public lighting: average spacing (between lamp posts) 4		less than 20 m	high quality	smaller spacing of permanent nighttime light sources in the segment: potential for significant maximization of adequate luminosity distribution in public spaces	Absence of luminaires or low lighting efficiency	Lowered luminaires for pedestrian circulation or high lighting efficiency
			from 20 to 35 m	medium-high quality	lower-medium spacing of permanent nighttime light sources in the segment: potential for relative maximization of adequate luminosity distribution in public spaces
			from 36 to 50 m	medium-low quality	medium spacing of permanent nighttime light sources in the segment: potential for relative minimization of adequate luminosity distribution in public spaces
			more than 50 m	low quality	greater spacing of permanent nighttime light sources in the segment: potential for significant minimization of adequate luminosity distribution in public spaces
Permanencies	Arrangement of seating and/or stopping furniture: average percentage		more than 50%	high quality	tendency for segment data located within the two upper quartiles: potential for substantial maximization of positive furniture arrangement attributes	Absence of benches and similar elements or in an exceptional state of discomfort (e.g., excessive spacing or insufficient quantity)	Benches and similar elements in an exceptional state of comfort and with universal accessibility solutions
			from 26 to 50%	medium-high quality	tendency for segment data located within the intermediate-upper quartile: potential for relative maximization of positive furniture arrangement attributes
			from 10 to 25%	medium-low quality	tendency for remaining segment data in the lower decile located within the lower quartile: potential for relative minimization of positive furniture arrangement attributes
			less than 10%	low quality	tendency for segment data located within the lower decil: potential for substantial minimization of positive furniture arrangement attributes
	Shading for seating and/or stopping: average percentage		more than 50%	high quality	tendency for segment data located within the two upper quartiles: potential for substantial maximization of positive shading attributes	Absence of shading components or in an exceptional state of discomfort	Shading components in an exceptional state of comfort
			from 26 to 50%	medium-high quality	tendency for segment data located within the intermediate-upper quartile: potential for relative maximization of positive shading attributes
			from 10 to 25%	medium-low quality	tendency for remaining segment data in the lower decile located within the lower quartile: potential for relative minimization of positive shading attributes
			less than 10%	low quality	tendency for segment data located within the lower decil: potential for substantial minimization of positive shading attributes
Socio-political
Institutional instruments/tactical tools		Maintenance of public spaces: average percentage (deteriorated spaces)	less than 10%	high quality	tendency for segment data located within the lower decile: indicative of smaller degree of maintenance precariousness in public spaces	Exceptionally negative state of conservation, inducing physical-social disorder	Exceptionally positive state of conservation, inducing physical-social order
			from 10 to 25%	medium-high quality	tendency for remaining segment data in the lower decile located within the low quartile: indicative of lower-median degree of maintenance precariousness in public spaces
			from 26 to 50%	medium-low quality	tendency for segment data located within the intermediate-low quartile: indicative of median degree of maintenance precariousness in public spaces
			more than 50%	low quality	tendency for segment data located within the two upper quartiles: indicative of greater degree of maintenance precariousness in public spaces
		Monitoring of public and private spaces: average number of resources (per 50 m) 5	more than 4 resources	high quality	greater availability of monitoring resources across public and private spaces along the segment: indicative of more significant spatial control	Absence of cameras or orientation of most cameras toward the interior of lots, and without constant police presence	Orientation of most cameras toward public spaces and with constant police presence
			from 3 to 4 resources	medium-high quality	medium availability of monitoring resources across public and private spaces along the segment: indicative of significant spatial control
			from 1 to 2 resources	medium-low quality	lower availability of monitoring resources across public and private spaces along the segment: indicative of insignificant spatial control
			nonexistence of resources	low quality	absence of monitoring resources across public and private spaces along the segment: indicative of lack of spatial control

Notes: ¹ = minimum adequate proportion sidewalk width/roadway width = 1/2 [71]; ² = minimum–adequate proportion block length/higher building height = 1/2 [72]; ³ = estimated mean density of Itumbiara = 24 inh./ha [61,62]; ⁴ = minimum adequate spacing of permanent nighttime light sources= 35 m [73]; ⁵ = minimum adequate spacing of control devices = 50 m (facial recognition) [72].

, the adopted classification system is presented. The measurement results of the parameters were grouped into four classes (high, medium-high, medium-low, and low quality), as products of the specificities of each variable, mostly expressed in average percentages or average proportions relative to the length of each segment or other measures. In certain situations, it was necessary to determine classification reducers and promoters, with the purpose of eventually lowering or raising a position, respectively.

The classification results were organized and tabulated in electronic spreadsheets of Excel for Windows, with the following values assigned per class: 4 for high; 3 for medium-high; 2 for medium-low; 1 for low. In the case of averages, the intervals corresponded to: 4.0 to 3.3 for high; 3.2 to 2.5 for medium-high; 2.4 to 1.7 for medium-low; 1.6 to 1.0 for low. After their joint interpretation in the form of an analytical synthesis, the findings were analyzed with the primary function of interpreting bases for management.

Statistical interpretations were conducted within the same Python environment, employing the Scientific Python (SciPy) library to estimate Spearman’s Correlation Coefficient (ρ) [68] across six spatial scenarios of Avenida Beira-Rio: three segments (North, Central, and South) versus two sides (Western and Eastern). A total of 15 parameters derived from socio-behavioral, socio-morphological, and socio-political variables were processed, with correlation magnitudes interpreted according to Cohen’s intervals: 0.10–0.29 (weak), 0.30–0.49 (moderate), and 0.50–1.00 (strong) [69].

Within this framework, an effort was made to connect socio-behavioral, socio-morphological, and socio-political variables with assumptions outlined in classical works and contemporary references on criminological theories. In doing so, the intention was to synthesize findings for an urban agenda aimed at integrated management of the urban landscape and public safety.

3. Results

In line with the previously detailed procedures, the findings are synthesized into analytical considerations regarding the selected district, the chosen street, and their respective segments.

3.1. Selecting the District

Table 3. A quantitative synthesis of the partial, total, and average classes of thefts and robberies by districts of Itumbiara in the before (2018–2019), parallel (2020–2021), and after (2022–2023) periods of COVID-19. Sources: [60,62].

Districts		Thefts									Total	Robberies									Total	Grand Total
		Before			Parallel			After				Before			Parallel			After
		2018	2019	Subtotal	2020	2021	Subtotal	2022	2023	Subtotal		2018	2019	Subtotal	2020	2021	Subtotal	2022	2023	Subtotal
1	Alto da Boa vista	62	33	95	43	19	62	29	43	72	229	22	24	46	9	9	18	9	5	14	78	307
23	Jardim América	34	33	67	38	19	57	22	27	49	173	19	26	45	5	6	11	8	0	8	64	237
50	Setor Afonso Pena	86	93	179	88	87	175	114	75	189	543	37	34	71	21	27	48	14	10	24	143	686
54	Setor Central	346	252	598	319	189	508	160	167	327	1.433	95	91	186	43	31	74	38	11	49	309	1.742
58	Setor Nossa Senhora da Saúde	42	31	73	42	26	68	22	35	57	198	15	11	26	3	11	14	11	2	13	53	251
61	Setor Novo Horizonte	44	56	100	41	20	61	28	45	73	234	15	10	25	16	11	27	7	3	10	62	296
63	Setor Paranaíba	28	29	57	24	17	41	14	13	27	125	17	10	27	6	10	16	3	5	8	51	176
64	Setor Planalto	43	36	79	26	29	55	36	37	73	207	20	20	40	8	9	17	9	3	12	69	276
67	Setor Rodoviário	49	38	87	46	23	69	29	17	46	202	20	16	36	9	14	23	3	3	6	65	267
70	Setor Santa Rita	27	42	69	30	41	71	25	31	56	196	19	13	32	21	16	37	13	6	19	88	284
71	Setor Santos Dumont	73	59	132	63	32	95	51	47	98	325	32	22	54	19	18	37	12	6	18	109	434
75	Setor Social	27	38	65	28	19	47	22	33	55	167	19	14	33	7	2	9	5	3	8	50	217
LEGEND
Thefts						Robberies
	low: <40 annual occurrences						low: <3 annual occurrences
	medium-low: 41–80 annual occurrences						medium-low: 3–6 annual occurrences
	medium-high: 81–120 annual occurrences						medium-high: 7–10 annual occurrences
	high: >120 annual occurrences						high: >10 annual occurrences
	no record						no record

Note: District numbering equivalent to that of Figure 1.

presents the 12 districts (14.4% of the total) with records of thefts and robberies above the averages in each year. From the interpretation of this information, the clear prominence of the Setor Central is inferred and adopted, therefore, as the specific study area. Its situation confirms that rates of these crimes are normally higher in city centers than in other regions of urbanized areas [74], although the city context plays a significant role in shaping overall criminal trajectories [75].

For thefts, the Friedman Test (

Table 4. The Friedman Test results (global) for thefts and robberies by districts of Itumbiara in the before (2018–2019), parallel (2020–2021), and after (2022–2023) periods of COVID-19. Sources: [60,62] (see Table 3).

Crimes	n	p Values	W de Kendall	Effect Magnitudes
Thefts	12	0.002	0.521	large
Robberies	12	<0.001	0.861	large

) revealed a significant overall difference across periods (p = 0.002; Kendall’s W = 0.52, large effect). Pairwise comparisons with Holm Correction showed significance between before (2018–2019) and during the pandemic (2020–2022 (adjusted p = 0.010), as well as between pre-pandemic and post-pandemic (adjusted p = 0.010), but not between the parallel and after COVID-19 periods (adjusted p = 0.556). This indicates that most of the decline in theft occurred during the pandemic and remained stable thereafter (

Table 5. The pairwise comparison results for thefts and robberies by districts of Itumbiara in the before (2018–2019), parallel (2020–2021), and after (2022–2023) periods of COVID-19. Sources: [60,62] (see Table 3).

Crimes	Comparison Periods	Adjusted p	r	Effect Magnitudes
Thefts	2018–2019 vs. 2020–2021	0.010	0.861	large
	2020–2021 vs. 2022–2023	0.556	0.181	small
	2018–2019 vs. 2022–2023	0.010	0.827	large
Robberies	2018–2019 vs. 2020–2021	0.008	0.816	large
	2020–2021 vs. 2022–2023	0.008	0.884	large
	2018–2019 vs. 2022–2023	0.008	0.884	large

). Regarding robberies, the Friedman Test also demonstrated a significant global difference across periods (p < 0.001; Kendall’s W = 0.86, large effect), with significance observed in all three pairwise comparisons after Holm Correction (adjusted p = 0.008), highlighting a consistent and progressive decrease in robbery throughout the three intervals.

As support for the pairwise comparisons above,

Table 6. The descriptive statistics results of the difference in case numbers of thefts and robberies by districts of Itumbiara in the before (2018–2019), parallel (2020–2021), and after (2022–2023) periods of COVID-19. Sources: [60,62] (see Table 3).

Crimes	Comparison Periods	n	Mean	SD	Median	Q1	Q3	Minimum	Maximum
Thefts	2018–2019 vs. 2020–2021	12	−24.3	24.5	−18.0	−34.0	−8.8	−90	2
	2020–2021 vs. 2022–2023	12	−39.9	74.0	−20.5	−31.0	−12.3	−271	10
	2018–2019 vs. 2022–2023	12	−15.6	53.8	−2.5	−14.3	10.5	−181	18
Robberies	2018–2019 vs. 2020–2021	12	−24.2	29.9	−20.0	−25.0	−11.8	−112	5
	2020–2021 vs. 2022–2023	12	−36.0	33.5	−29.0	−36.3	−18.0	−137	−13
	2018–2019 vs. 2022–2023	12	−11.8	9.0	−12.5	−18.3	−3.8	−25	−1

Notes: n = sample size (number of observations in the sample). SD = standard deviation (measure of data variability). Median = value at which 50% of the observations fall below in the variable. Q1 = value at which 25% of the observations fall below in the variable. Q3 = value at which 75% of the observations fall below in the variable. Min = minimum recorded value in the sample. Max = maximum recorded value in the sample.

presents the descriptive statistics across the 12 districts, reflecting the difference in case numbers between each compared period (e.g., thefts during the pandemic minus thefts before, calculated district by district). A negative difference indicates a reduction in cases within the districts across the two periods.

The mean of the differences is strongly influenced by the Central Sector (the district with a volume often much larger than the others). For instance, the average reduction in thefts (−39.9) between the pre- (2018–2019) and post-pandemic (2022–2023) periods is considerably greater than the median (−20.5), precisely due to this neighborhood. The median, being less sensitive to this extreme value, represents more accurately the typical reduction observed in a district: approximately 18 fewer thefts and 20 fewer robberies between the before (2018–2019) and parallel pandemic (2020–2021) periods, with similar or greater reductions between before and after the pandemic, consistent with the statistical significance found in the tests.

Situated in the south-center portion of the urban perimeter of Itumbiara (Figure 2), the selected district—Setor Central—covers an area of about 1.35 km². In 2025, it housed about 3300 residents [61,62]. To enable the interpretation of results within a specific location, the next step involves the identification of the street for detailed examination.

3.2. Selecting the Street

The mapping of thefts in the Setor Central (Figure 3) identifies Avenida Beira-Rio as the roadway with the highest concentration of this offense throughout all examined periods. The cartographic depiction of robberies (Figure 4) likewise designates the same street as the site of the most notable incidents within analytical intervals. These patterns are further reinforced by the summary maps shown in Figure 5.

The Getis-Ord Gi\* statistic results indicates that in all the years analyzed (2018 to 2023) and across the entire period, between 1.6% and 3.1% of the streets were identified as statistically significant clusters (hot spots) at the 95% or 99% confidence levels, consistently concentrated in the same few road axes of the sector, with p < 0.001 in the vast majority of cases. Therefore, thefts and robberies are spatially concentrated in specific streets of the Central Sector, as identified for the spatial autocorrelation (

Table 7. A synthesis of the hot spot analysis (Getis-Ord Gi\* statistic) of thefts and robberies by streets of the Setor Central of Itumbiara in the before (2018–2019), parallel (2020–2021), and after (2022–2023) periods of COVID-19. Sources: [60,62,63] (see Figure 3, Figure 4 and Figure 5).

Crime	Year	n (Streets)	Significant Hot Spots	% Hot Spots	More Significant Streets	Maximum Gi-Z	Minimum p
Thefts	2018	64	1	1.6	Avenida Beira-Rio	5.54	<0.001
	2019	64	1	1.6	Avenida Beira-Rio	5.14	<0.001
	2020	64	2	3.1	Avenida Beira-Rio Avenida Paranaíba	6.21 3.60	<0.001
	2021	64	2	3.1	Avenida Beira-Rio Avenida Paranaíba	4.48 3.20	<0.001
	2022	64	2	3.1	Avenida Beira-Rio	5.47	<0.001
	2023	64	1	1.6	Avenida Beira-Rio	6.44	<0.001
	Total	64	2	3.1	Avenida Beira-Rio	5.29	<0.001
Robberies	2018	42	1	2.4	Avenida Beira-Rio	4.2	<0.001
	2019	42	1	2.4	Avenida Beira-Rio	5.66	<0.001
	2020	42	1	2.4	Avenida Beira-Rio	4.54	<0.001
	2021	42	1	2.4	Avenida Beira-Rio	5.09	<0.001
	2022	42	1	2.4	Avenida Beira-Rio	4.61	<0.001
	2023	42	1	2.4	Avenida Beira-Rio	4.42	<0.001
	Total	42	1	2.4	Avenida Beira-Rio	5.55	<0.001

), with particular emphasis on Avenida Beira-Rio.

From 2018 to 2023, theft and robbery records along Avenida Beira-Rio amounted to 146 active points categorized as ‘public place’ and ‘street/avenue’. The former accounted for 95 (65.1%), while the latter comprised 51 (34.9%). A gradual decline in both figures and proportions is further observed across each analytical period (Figure 6), registering a substantial reduction of 87.8% in the overall criminal occurrences examined throughout the six years.

The marked decline in thefts and robberies during the pandemic period can be explained by restrictions imposed for social isolation due to the disease. However, the persistence of similar or even lower levels in 2022 and 2023 may be linked to perceptions of behavioral change at both individual and collective scales [76].

The monthly distribution of offenses (Figure 7) does not reveal any significant pattern, meaning it fails to distinguish differentiated conditions for classifying analytical variables. Human activities in public spaces are often associated with seasonal factors and non-school periods. In this regard, it is worth noting that Itumbiara’s climatic characteristics are relatively stable, with average monthly maximum and minimum temperatures of 33 °C and 22 °C, respectively, even though rainfall varies from 2.0 mm to 281 mm, with heavier precipitation in summer [77]. Furthermore, there is evidence that hotter cities show a weaker correlation between climate and crime [78].

Based on the statistical assessment of Spearman’s Coefficient (ρ) [68] (0.186; p-value 0.600—weak), no significant correlation emerges between months of the year and thefts and robberies. This outcome suggests that seasonal drivers of thefts in June and October, for instance, do not necessarily align with periods of increased robberies. Monthly crime distributions appear to operate under distinct logics of opportunity for each modality [9,10].

Prior to the pandemic, there was a tendency for criminal activity to occur during nighttime hours (Figure 8). This inclination persisted in the first year (2018) of the pandemic period, while the second (2019) showed an unspecific shift.

The tendency for nighttime criminal activity reappears in the first year (2022) after the pandemic, remaining only for robberies in the following year (2023). Even so, there are sufficient indications to consider nocturnal conditions in the classification of analytical variables. This situation is common in areas of ‘night-time economy,’ where potential victims may be intoxicated or concentrated within a small geographic space [79]. After dusk, the external environment changes markedly, which can increase criminal opportunities—though these could be mitigated through appropriate public lighting solutions [80].

Through the evaluation of Spearman’s Coefficient (ρ) [68] (0.809; p-value 0.015—strong), unlike what was observed for months, a positive and statistically significant correlation is found for daily intervals. This indicates that the temporal dynamics of thefts and robberies are highly similar in the Setor Central, with periods of higher incidence of one crime (particularly at night, from 18:00 to 23:59) tending to coincide with elevated levels of the other. Such a condition underscores the relevance of locational aspects and public lighting [21,22,23,80]. Subsequently, the closest interpretation of the path is sought.

3.3. Segmenting the Route

Avenida Beira-Rio takes its name from bordering the Paranaíba River, which marks the boundary between the states of Goiás and Minas Gerais (Figure 9). Within its about 750 m stretch inside the Setor Central, it is defined as a roadway axis with a central median strip and lanes in opposite directions. From a morpho-urban perspective, it is divided into three main segments (Figure 10): Northern—from Avenida Trindade to Avenida Paranaíba, without intersections; Central—from the latter to Rua Damores Amaral Medeiros, intersected by Rua Dr. Valdino Vaz and Rua Sebastião Caldino; and Southern—from this point to Avenida Rogelina Maria de Jesus, crossed diagonally by Rua Antônio Marques. The Eastern Side of each portion borders the river’s permanent preservation area, in part arranged as a leisure park.

The Northern Segment (350 m in length) (Figure 11) features palm trees along much of its central median strip. Its Western Side is largely composed of spacious plots, marked by notable urban voids. In the Central Segment (280 m), the same type of lane divider continues, with identical Palmaceae species. Its Western portion contains plots displaying varied levels of building occupation and mixed functions. The Southern Segment (125 m) lacks vegetation in the central median strip. On the West Side, in the mid-northern area, it is bordered by a densely wooded square. In the mid-southern zone, there is a substantial urban void. Each side of the street segments is assessed according to the previously specified socio-behavioral, socio-morphological, and socio-political variables (see Section 2.3).

3.4. Evaluating Variables

The morpho-urban conditions across the three street segments establish distinct situations for classifying the analytical variables. For socio-behavioral variables (

Table 8. An evaluation panel of the analytical variables and respective parameters of the segments of Avenida Beira-Rio in the Setor Central of Itumbiara. Source: Adopted methodological procedures (see Section 2.3 and Appendix A).

Variables		Parameters			Northern Segment			Central Segment			Southern Segment			Overall Averages
					Western Side	Eastern Side	Averages	Western Side	Eastern Side	Averages	Western Side	Eastern Side	Averages
Socio-Behavioral Variables
Evasion		Escape routes			3 R1	1	2.0	1	1	1.0	1	1	1.0	1.3
Concealment		Physical barriers			4	3 R2	3.5	4	3 R2	3.5	2	1	1.5	2.8
Shielding		Visual obstructions			4	1	2.5	4	1	2.5	2	1	1.5	2.2
Overall Averages					3.7	1.7	2.7	3.0	1.7	2.3	1.7	1.0	1.3	2.1
Socio-Morphological Variables
Dimensions		Street scale			1	4 P1	2.5	3	4 P1	3.5	4 P2	4 P1	4.0	3.3
		Urban scale			4	4	4.0	4	4	4.0	4	4	4.0	4.0
Densities		Constructive/ population densification			1	1	1.0	4	1	2.5	1	1	1.0	1.5
		Diversity of uses			2	1	1.5	4	1	2.5	1	2	1.5	1.8
Accessibilities		Relations between public and private space			1	4	2.5	1	4	2.5	4	4	4.0	3.0
		Spatial intervisibility			1	4	2.5	2	4	3.0	4	4	4.0	3.2
Attractiveness		Local references			1	1	1.0	1	1	1.0	1	1	1.0	1.0
		Public lighting			2 R3	4 P3	3.0	2 R3	4 P3	3.0	2 R3	4 P3	3.0	3.0
Permanencies		Arrangement of seating and/or stopping furniture			1	3 R4	2.0	1	3 R4	2.0	3 R5	1	2.0	2.0
		Shading for seating and/or stopping			1	3	2.0	1	3	2.0	4	4	4.0	2.7
Overall Averages					1.5	2.9	2.2	2.3	2.9	2.6	2.8	2.9	2.9	2.6
Socio-Political Variables
Institutional instruments/ tactical tools		Maintenance of public and private spaces			1	4	2.5	2	4	3.0	3	4	3.5	3.0
		Monitoring of public and private spaces			1	1	1.0	2 P4	1	1.5	1	1	1.0	1.2
Overall Averages					1.0	2.5	1.8	2.5	2.0	2.0	2.0	2.5	2.3	2.1
LEGEND
	high quality			medium-high quality				medium-low quality				low quality

Notes: ^P1 = application of promoter due to the existence of a promenade with universal accessibility solutions. ^P2 = application of promoter due to the expansion of sidewalks into open areas. ^P3 = application of promoter due to the presence of high-efficiency lighting luminaries. ^P4 = application of promoter due to the potential for natural surveillance through the presence of people. ^R1 = application of reducer due to the possibility of escape through wall crossing. ^R2 = application of reducer due to the possibility of concealment caused by the drop on the riverbank. ^R3 = application of reducer due to the existence of public lighting only along the median strip. ^R4 = application of reducer due to the excessive distance between benches. ^R5 = application of reducer due to the inadequate quantity of benches.

), escape routes show the lowest overall performance (1.3—low class), indicating a high potential for evasion along the roadway, which reduces the likelihood of offenders being apprehended. The highest average is attributed to physical barriers (2.8—medium-high class), revealing that the limited presence of obstacles makes potential targets more visible to other passersby, thereby increasing opportunities for natural surveillance and decreasing the chances of shielding the aggressor [5,9,82]. The Southern Segment is the most vulnerable to theft and robbery (overall average of 1.3—low class), with inferior performance across all evaluated parameters. The most favorable condition is recorded in the Northern Segment (overall average of 2.7—medium-high class), mainly due to the negligible presence of physical barriers that could conceal criminal activity. Without the adoption of promoters, classification reducers were applied in cases of possible escape through the crossing of the extensive wall on the Western Side of the Northern Segment, due to the idleness of the land plot, as well as concealment of offenders by the topographic slope of the riverbank throughout the first two segments.

For the socio-morphological variables, the lowest rating corresponds to local references (1.0—low class), followed by constructive population density (1.5—low class), highlighting problems of attractiveness, which may bring victims and aggressors closer through generated opportunities [10,31,32,83]. The best performance is observed for the urban scale (3.8—high class), confirming the importance of certain spatial–environmental metrics for appropriate relationships between the built environment and criminal occurrence, given their influence on criminal behaviors [15,37,84].

The Southern Segment is the least prone to theft and robbery (overall average of 2.9—medium-high class), with good ratings across all parameters, except for constructive population density, diversity of uses, local references, and arrangement of seating and/or stopping furniture. The poorest condition is observed in the Northern Segment (overall average of 2.2—medium-low class), resulting from nearly all attributes, except street and urban scales, relations between public and private space, spatial intervisibility, and public lighting.

The application of classification promoters was justified by the existence of a pedestrian promenade with universal accessibility solutions, the expansion of sidewalks into open areas, and the presence of high-efficiency lighting fixtures. Conversely, reducers were applied in cases of the existence of public lighting only along the median strip, the excessive distance between benches, and their inadequate quantity in some spaces.

For the two parameters of the socio-political variables evaluated, the lowest overall performance (1.2—low class) pertains to the monitoring of public and private spaces, revealing that the absence of devices for this purpose facilitates the occurrence of crimes [85]. In contrast, spatial maintenance is relatively satisfactory (3.0—medium-high class), indicating its positive influence on reducing criminality [40,50,51,52]. The only classification promoter is applied to the Western Side of the Central Segment, regarding its potential for natural surveillance through the presence of people [9].

By integrating the set of variables (

Table 9. A general evaluation panel of the analytical variables and respective parameters of the three segments of Avenida Beira-Rio in the Setor Central of Itumbiara. Source: Adopted methodological procedures (see Section 2.3 and Appendix A).

Variables				Northern Segment				Central Segment				Southern Segment					Overall Averages
				West Side	East Side	Averages		West Side		East Side	Averages	West Side		East Side		Averages
Socio-Behavioral				3.7	1.7	2.7		3.0		1.7	2.3	1.7		1.0		1.3	2.1
Socio-Morphological				1.5	2.9	2.2		2.3		2.9	2.6	2.8		2.9		2.9	2.6
Socio-Political				1.0	2.5	1.8		2.0		2.5	2.3	2.0		2.5		2.3	2.1
Overall Averages				2.1	2.4	2.2		2.4		2.4	2.4	2.2		2.1		2.1	2.2
LEGEND
	high quality		medium-high quality						medium-low quality						low quality

), the lowest overall performance is found in the socio-political dimension, followed by the socio-behavioral (2.1 for both—medium-low class). The socio-morphological variables achieve a slightly higher score (2.6—medium-high class). In summary, all Avenida Beira-Rio segments fall within the lower median condition (values ranging from 2.1 to 2.4), with an overall average of 2.2.

Considering the total of 15 parameters derived from socio-behavioral, socio-morphological, and socio-political variables, the correlation matrix constructed through Spearman’s Coefficient (ρ) [68] (

Table 10. A correlation matrix of the analytical variables and respective parameters for the sides of the three segments of Avenida Beira-Rio in the Central Sector of Itumbiara. Source: Adopted methodological procedures (see Table 8 and Table 9).

Segments	North- Western	North- Eastern	Central- Western	Central- Eastern	South- Western	South- Eastern
North- Western	1.00	−0.15	0.51	−0.15	−0.12	−0.17
North- Eastern	−0.15	1.00	−0.07	1.00	0.85	0.85
Central- Western	0.51	−0.07	1.00	−0.07	−0.07	−0.07
Central- Eastern	−0.15	1.00	−0.07	1.00	0.85	0.85
South- Western	−0.12	0.85	−0.07	0.85	1.00	0.80
South- Eastern	−0.17	0.85	−0.07	0.85	0.80	1.00

) reveals distinct patterns between the two sides of Avenida Beira-Rio. A perfect positive correlation (1.00) is observed between the Eastern Side of the Northern and Central segments. This occurs because both share a substantial portion of the parameters, presenting elevated levels (class 4—high) of public lighting, spatial intervisibility, and maintenance of public spaces, but low values (class 1—low) of diversity of uses and constructive population.

There is also notable coherence in the Southern Segment, with its Western and Eastern sides exhibiting a strong positive correlation (0.80). Unlike the others, it maintains uniformity in variables such as shading and urban scale (class 4—high on both flanks), with a more homogeneous morphology at this end of the street due to the presence of tree masses (square and park). Likewise, similarity is observed between the Eastern sides of the other two segments and the South, with remarkably high correlations, further highlighting the positive natural features of the urban landscape.

The Western sides of the Northern and Central segments display negative or insignificant correlations with almost all other relationships, except between themselves (0.51). Although more built-up scenarios, they reveal less favorable indices for evasion (particularly escape routes in the Northwestern Side—class 3—medium-low) and shielding (visual obstructions in the Northwestern Side—class 4—high), contrasting sharply with the East.

In summary, this statistical analysis demonstrates the spatial duality of the Eastern Side, strongly correlating with the Southern segment, whereas the Western sides of the Northern and Central segments exhibit morphological conditions that favor concealment and visual isolation, which may result in direct implications for integrated public security management.

Confirming the street as a platform for public life [86,87], these results support the diagnosis that theft and robbery occurrences align with factors facilitating criminal behavior, linked both to landscape features that intensify crime opportunities and to inefficient institutional instruments and tactical tools of public management. Thus, it becomes possible to discuss and organize foundations for structuring urbanistic-criminal policies relevant to the theme. In this regard, they should aim at integrating measures from different sectors.

4. Discussion

Based on the case studied, and for the shared management of a safe landscape, several guidelines can be highlighted for an urban agenda based on the analytical variables. In this context, the recommendations can be linked to both classical and contemporary criminological theories, as synthesized in

Table 11. A synthesis panel of guidelines for the urban agenda based on analytical variables associated with criminological theories. Sources: Listed in the cells.

Criminological Theories 1 (Classical Works and Contemporary References)	Assumptions	Analytical Variables
		Socio-Behavioral (Evasion, Concealment, Shielding)	Socio-Morphological (Dimensions, Densities, Accessibilities, Attractiveness, Permanencies)	Socio-Political (Institutional Instruments/Tactical Tools)
Routine Activities [5,6,7,8]	Crime resulting from the interaction between the existence of a target, the absence of a guardian, and the presence of an offender.	Insufficient alternatives for potential aggressors and removal of physical and visual barriers.	Reduction in unsafe practices, accessibility to escape routes, and attractiveness for offenders, together with encouragement of permanence for guardians.	Formulation of guidelines and implementation of protective measures for targets, attraction of guardians, and suppression of offenders.
Criminal Opportunities [9,10]	Limited disposition of citizens to maintain a crime-prevention posture.	Attention of spatial users to escape routes and physical and visual obstructions.	Attention of spatial users to unsafe dimensions, densities, accessibility, attractiveness, and permanencies.	Guidance for minimizing criminal opportunities.
Marginal Labeling [11,12,13]	Transformation of the individual into what they are labeled.	Reduction in social discrimination and consequent decrease in offenders’ propensity to flee, hide, and conceal.	Social decriminalization through dimensions, densities, accessibility, attractiveness, and spatial permanencies.	Awareness and implementation of actions for reducing socio-criminological discrimination.
Criminal Economy [11,14]	Evaluation of costs and gains of criminal actions.	Insufficiency of alternative gains from evasion, concealment, and hiding.	Reduction in criminal gains through dimensioning, densities, accessibility, attractiveness, and permanencies.	Construction and institution of alternatives for reducing criminal gains.
Rational Choice [5,6,11,15,16,17,18]	Option of committing crime based on weighing benefits against punishment.	Reduction in spatial advantages for evasion, concealment, and hiding.	Attenuation of criminal advantages through dimensioning, densities, accessibility, attractiveness, and permanencies.	Outline and implementation of spatial disadvantages for offenders.
Social Disorganization [19,20]	Probability of criminal occurrence in communities collapsed in customs and social opportunities.	Community strengthening for prevention of evasion and monitoring of physical and visual barriers.	Community interaction in defining dimensions, densities, accessibility, attractiveness, and permanencies.	Design and institution of community strengthening initiatives for crime prevention.
Locational Criminality [21,22,23]	Criminal actions linked to opportunities favored by the physical environment.	Control of escape routes, hiding places, and concealment areas for offenders.	Establishment of measures regarding dimensions, densities, accessibility, attractiveness, and permanencies to discourage crime.	Design and implementation of principles for public spaces guided by safety and adapted to local specificities.
Personal Space [24,25,26]	Personal physical environment influencing the propensity for criminal behavior.	Personal space restricting escape routes and mitigating physical and visual barriers.	Dimensions, densities, access, attractions, and permanencies related to personal space.	Guidance and implementation of measures to minimize criminal opportunities associated with safe personal spaces.
Social Cohesion [27,28]	Urban permanencies that strengthen solidarity but also create opportunities for criminal activities.	Separation of urban permanencies from escape routes, avoiding crowding that generates physical and visual obstructions.	Separation of urban permanencies from escape routes, ensuring accessibility, attractiveness, and safety, without crowding that produces physical and visual barriers.	Design of principles for public spaces guided by solidarity and safety, with permanencies that are both supportive and secure.
Territorial Defense [29,30]	The importance of the surrounding environment for spatial maintenance and social responsibility.	Social responsibility in hindering escape routes and monitoring physical and visual barriers.	Social responsibility in defining dimensions, densities, accessibility, attractiveness, and permanencies.	Awareness and assurance of social responsibility in crime monitoring.
Social Density [31,32]	Constructive population rates generate more interactive behaviors, with environmental and behavioral regulation.	Population interaction in monitoring escape routes and physical or visual obstructions.	Population interaction in defining dimensions, densities, access, attractions, and urban permanencies.	Orientation and assurance of population interaction in crime monitoring.
Safe Spaces [9,24,25,33]	Environments with easy access and mobility inducing the presence of ‘natural guardians.’	Safety conditions impeding escape routes and mitigating physical and visual barriers.	Safety conditions associated with dimensions, densities, access, attractions, and permanencies.	Conception and implementation of guidelines for the design of safe spaces.
Environmental Design [15,34]	Organization of physical space for crime prevention and control.	Minimization of spatial opportunities for evasion, concealment, and hiding of offenders.	Enhancement of measures, densities, access, attractions, and appropriate permanencies for safe landscapes.	Design and implementation of measures to minimize spatial opportunities for the occurrence of crimes.
Universal Design [35,36]	Spatial accessibility for all people.	Accessible design solutions, with control of escape routes and places for concealment and hiding of offenders.	Dimensioning for accessible design solutions, including in areas of density, with universal accessibility, safe attractions, and permanencies.	Design and implementation of accessible project solutions for crime control.
Defensible Space [34,37,38,39]	Spatial design with a security and defense function conducted by citizens themselves.	Spatial design hindering escape routes and mitigating physical and visual barriers.	Design with dimensions, densities, accessibility, attractiveness, and permanencies for spatial defense.	Conception and implementation of principles for the design of defensible public spaces.
Spatial Needs [24,25,40]	Spatial construction influencing users’ sense of security.	Spatial limitation of escape routes and of physical and visual barriers.	Reduction in unsafe measures for potential targets and of accessibility to offenders’ escape routes, with favoring attractiveness and permanencies for guardians.	Outline and implementation of spatial limitation options for criminal activities.
Spatial Victimization [24,25,41]	Places of social interaction for longer periods facilitating the existence of targets.	Distancing potential victims from escape routes and physical and visual barriers.	Distancing, decongestion, and inaccessibility for potential victims, with reduction in attractions and permanencies contributing to their insecurity.	Guidance and implementation of measures to minimize chances of victimization in public spaces.
Social Defense [42,43]	Public security as a responsibility not only of the state.	Social participation in monitoring escape routes and physical or visual obstructions.	Social participation in monitoring dimensions, densities, accessibility, attractiveness, and permanencies.	Organization and encouragement of social involvement in crime monitoring.
Social Control [44,45,46,47,48]	Society as responsible for its own stability, with clear rules for crime prevention.	Transparent dissemination of penalties for evasion and establishment of norms preventing the creation of physical and visual barriers.	Regulation of dimensions, densities, access, and attractions related to permanencies and circulations.	Construction of social norms for crime prevention and their formal institution for criminal repression.
Natural Surveillance [9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49]	People as ‘guardians’ of public areas.	Guardians’ attention to escape routes and physical and visual barriers.	Guardians’ attention to dimensions, densities, accessibility, attractiveness, and spatial permanencies.	Construction and institution of alternatives for attracting spatial guardians.
Broken Windows [50,51,52]	Physical disorder enhancing criminality.	Physical ordering to hinder escapes and mitigate physical and visual barriers.	Physical ordering to adjust dimensions, densities, accessibility, attractiveness, and permanencies for spatial defense.	Conception and implementation of principles of physical–spatial ordering for crime prevention.
Results-Based Management [53,54]	Systematic performance analysis of security, with evaluation of strategic and tactical actions.	Evaluation of strategic and tactical actions for containment of escapes and elimination of physical and visual obstructions.	Evaluation of strategies and tactics for dimensioning, densities, access, attractions, and permanencies.	Evaluation of strategic instruments and tactical tools for crime containment.
Criminal Mapping [24,25,55]	Data visualization on crimes supporting security management.	Visualization in maps of potential escape routes and physical and visual barriers.	Simulation in maps of dimensions, densities, access, attractions, and safe permanencies.	Simulation in maps of potential security strategies and tactics.

Note: ¹ = theories presented following the same sequence outlined in Section 1 (Introduction).

. Following the same sequence established in Section 1 (Introduction), where theoretical postulates were arranged according to their primary emphasis on socio-behavioral issues—principally associated with offenders’ actions and shaped by collective customs; socio-morphological aspects—denoting attributes of urban form that either facilitate or restrict unlawful practices; and socio-political dimensions—connected to mechanisms of municipal administration and civic participation, the synthesis panel of guidelines highlights the interrelations between the theories discussed and all analytical variables employed in the research undertaken.

The examination of the content of the synthesis panel of guidelines for the urban agenda elucidates that many criminal solutions address multiple theoretical postulates, while at the same time there is simultaneity in the alignment of principles with different analytical variables. These conditions reveal, on one side, the multidisciplinary nature of criminology [88], and on the other, the intricacy of urbanism [89].

4.1. Key Theoretical–Practical Contributions

In terms of socio-behavioral theories, emphasis should be placed on controlling evasion by minimizing factors that enable escape routes for offenders and hinder police pursuit, such as short blocks, proximity to corners, and other elements that define dispersal paths and facilitate concealment, restricting apprehension. Likewise, progressive elimination of components that create physical or visual hiding places for aggressors is essential, as these enable surprise attacks and limit victims’ reaction [5,9,82].

Except, at least partially, in relation to physical barriers, what emerges from the case study in Itumbiara is almost the reverse of the prescribed conditions, thereby substantiating, in particular, a considerable share of the approaches advanced by theories of Routine Activities [5,6,7,8], Criminal Opportunities [9,10], Criminal Economy [11,14], and Rational Choice [5,6,11,15,16,17,18], since existing conditions for the convergence of available targets, absence of guardianship, and criminal presence amplify both the limited willingness of citizens to adopt preventive behavior and the offenders’ evaluation of costs and benefits prior to acting, with perpetrator decisions ultimately shaped by perceived advantages weighed against eventual sanctions. It should be emphasized that incorporating additional parameters could further strengthen perspectives associated with Marginal Labeling [11,12,13] and Social Disorganization [19,20], avoiding the transformation of individuals into identities imposed by categorization and preventing crime within communities weakened by eroded customs and diminished social opportunities.

Regarding socio-morphological theories, guidelines include adjustments to dimensions, both at the street scale—with sidewalks that facilitate the circulation of spatial guardians—and at the urban scale, through built elements that attract and engage users, fostering natural surveillance [15,16,17,37,38,84]. Density should be encouraged by promoting constructive population concentration, increasing the number of pedestrians in outdoor areas, and diversifying land uses to enhance vitality and safety [32,33,83].

The case of Itumbiara reveals, on one hand, favorable conditions regarding street and urban scales, in contrast to constructive population densification and diversity of uses. This situation can be explained both by the medium size of the city and by the presence of green areas along Avenida Beira Rio. Nevertheless, it can be aligned with principles of Territorial Defense [29,30], Social Density [31,32], and Safe Spaces [9,24,25,33], since the surroundings exert a decisive influence on sustaining spatial order and promoting social responsibility, with densification levels fostering interactive conduct through environmental and behavioral regulation, while accessible settings encourage the presence of ‘natural guardians’, strengthening collective safety.

Accessibility improvements are also necessary, strengthening connections between public and private spaces, facilitating movement, and promoting social interaction. Spatial intervisibility should be enhanced by encouraging transparent façades and openings that allow visual contact for monitoring [87]. Attractiveness can be reinforced through distinctive places and landmarks that draw diverse groups, as well as through improved public lighting, with adequate spacing of permanent light sources to ensure quality illumination [79,80]. Permanence should be supported by providing seating and resting furniture, combined with appropriate shading, to promote environmental comfort and encourage people to remain in public spaces [32,33].

For this set of parameters concerning accessibility, attractiveness, and permanencies, the example of the Brazilian city yields more suitable outcomes for most of them, reinforcing statements from theories on Locational Criminality [21,22,23], Personal Space [24,25,26], Social Cohesion [27,28], and Defensible Space [34,37,38,39], since criminal dynamics arise from opportunities shaped by the physical environment, where individual surroundings influence behavioral propensity, public areas occupancy foster solidarity yet open avenues for unlawful acts, and urban design incorporates security and defense functions enacted by citizens themselves.

Less favorable conditions are identified along Avenida Beira-Rio in relation to the arrangement of seating and/or stopping furniture, as well as local references, underscoring the importance of observation to certain theoretical contributions, including those linked to principles of Spatial Needs [24,25,40].

Undeniably, these socio-morphological issues are supported by Environmental Design [15,34], with the organization of physical space aimed at crime prevention and control. However, along Avenida Beira-Rio, solutions of Universal Design [35,36] are scarce, virtually without inclusive design that ensures safe places. It is also worth noting the inherent ambiguity of some theories, since several advocate for the presence of individuals as ‘natural guardians,’ whereas Spatial Victimization [24,25,41], for instance, suggests that locals of prolonged social interaction function as factors increasing target availability for crime.

In relation to socio-political theories, attention should be directed toward improving institutional instruments and tactical tools for maintaining and monitoring public and private spaces. This includes evaluating the condition of sidewalks, roadways, and other components that may compromise physical–territorial order [40,50,51,52]. Special focus should be given to the presence of permanent personnel and spatial control devices [85]. Nevertheless, these latter components exhibit significant evaluative constraints in Itumbiara, particularly along Avenida Beira-Rio, where, at the time of the empirical surveys, no elements of monitoring of public and private spaces were identified, even when considering security as a responsibility beyond the state, as established by perspectives of Social Defense [42,43] and Social Control [44,45,46,47,48].

Moreover, it deserves particular mention that thefts and robberies frequently occur at night in Itumbiara, which undermines, especially under conditions of low illumination, the theoretical framework of Natural Surveillance [9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49], given the tendency toward a decline in the presence of people in public areas during this period. Even for the Brazilian city under study, the systematization of performance analysis of security is advocated, with assessment of strategic and tactical actions, from the perspective of Results-Based Management [53,54], supported by Criminal Mapping [24,25,55]. In this way, the deleterious effects of physical disorder enhancing criminality can be prevented, as explored by the Broken Windows theory [50,51,52].

In this sense, the critical discussion reinforces that urban policies must transcend reactive models and adopt preventive strategies that integrate public security, urban planning, and social participation. For this, master plans, urban codes, tactical instruments, and policing practices must be restructured in an articulated manner—a condition identified as essential for resilient and safe cities.

In this context, the main contributions of the research lie in establishing an integrative framework that connects public security, urban landscape, and institutional management, while providing empirical evidence that underscore systemic vulnerabilities across spatial, operational, and political dimensions. It demonstrates that criminal dynamics are structurally embedded rather than merely statistical, with temporal analysis confirming persistence of thefts and robberies despite pandemic fluctuations.

Despite the relevant limitations outlined below, the investigation also advances criminological theories by showing how spatially constructed opportunities interact with behavioral variables; consolidates a replicable methodological model that merges morphological analysis, ordinal classification, and integrated parameters assessment; and ultimately offers both theoretical reinforcement and practical guidance for designing resilient and preventive urban policies in contemporary contexts. The analysis of the study area shows that the criminal patterns observed reflect structural phenomena that go beyond mere numerical incidence of crimes. They reveal systemic fragilities combining physical vulnerabilities, operational restrictions, and political–institutional gaps.

4.2. Main Conceptual–Empirical Limitations

The systematization of the presented theories, while not encompassing all possible perspectives on the relationship between crime and urbanism, serves solely to support the intended analyses of this research. For other investigative lens, these frameworks require reformulation, with the potential adoption of more specific conceptual approaches tailored to each reality under interpretation. Moreover, theoretical contributions from diverse fields of knowledge could enhance outcomes for effective integrated management of the urban landscape and public security.

In empirical terms, although the preliminary formulation of a methodological essay was intended, with some contributions noted above, the most crucial limitation of the investigation lies in the fact that the case analysis encompasses only an urban district and three distinct segments (Northern, Central, Southern) of a single street (Avenida Beira-Rio) in the examined city (Itumbiara), with a reduced number of criminal incidents. The reliance on this micro-spatial sample does not allow for the generalization of findings to support a complex indicator system or the in-depth reviews of various theories.

Furthermore, this empirical restriction prevents certain specific statistical assessments, valid only for larger samples. It must also be noted that these research limitations constrain the generalizability of the findings to other urbanized contexts, making it advisable for future studies to expand the geographical scope of analyses—both to additional streets in the Central Sector and other road axes across the remaining districts of Itumbiara, as well as to other cities—in order to progressively refine this methodological essay.

The temporal scope, limited to three defined periods around the COVID-19 pandemic, may not capture longer-term fluctuations in criminal dynamics or broader socio-spatial transformations. The analytical framework, although innovative, depends on ordinal classifications that can oversimplify complex interactions among socio-behavioral, socio-morphological, and socio-political variables. Furthermore, the data set is restricted to theft and robbery records, leaving out other forms of crime that could influence urban vulnerabilities.

The use of four-level ordinal scales for assessing analytical parameters is justified to capture upward or downward tendencies in the results, although a higher even number of classes could allow greater analytical discrimination. In this way, numerical and/or percentage thresholds might also become more precise. It should be noted that certain differences in categorization may influence the robustness of the analytical outcomes, particularly in cases where elements could be measured with greater accuracy within the range of values of the local reality itself. Moreover, another possibility for future research concerns the assignment of weights according to the significance of each parameter, aiming at further refinement of the evaluative results, as an enhancement of the proposed methodological essay itself.

As the overall quality of the socio-morphological, socio-behavioral, and socio-political parameters in the study area is classified as ‘medium-low’ and the reality of the studied city does not reveal significant changes in socio-morphological (structural/physical) or socio-political (managerial/institutional) variables, it can be inferred that socio-behavioral dimensions would require methodological complements beyond metric-locational approaches to escape routes and physical/visual barriers, commonly examined in environmental criminology, toward perspectives more directly related to the assessment of human behavior itself.

Institutional and policy-related insights are derived from secondary sources, which may not fully reflect the nuances of local governance practices. In the case of public security, there are also issues related to sensitive information, whose access is restricted by legal data protection norms. Furthermore, crime under-reporting and inadequate recording are acknowledged, representing significant challenges to be addressed in the evaluation of criminality in Brazilian cities.

Therefore, it is reiterated that the experiment should be regarded as a preliminary methodological essay, whose future directions ought to include expanding the sample across diverse cities scales and geographic positions. They should also encompass adapting the analytical variables to each local reality, as well as the occasional removal of parameters or inclusion of new ones. In summary, the set of general guidelines identified provides a base for the concluding considerations of this debate.

5. Conclusions

Attesting to the achievement of the objective of analyzing public spaces in Itumbiara as inputs for integrated management policies of safe landscapes, responses are presented to the investigative question regarding the identifiable foundations for sharing urbanistic-criminal solutions within the framework of an urban agenda. At first, the answers refer to analytical results of districts according to theft and robbery incidence in the defined periods (before—2018–2019; parallel—2020–2021; and post-pandemic—2022–2023). Based on the discrimination of the study unit, pertinent to the Setor Central due to its higher crime occurrence, the road axis of Avenida Beira-Rio within this administrative region is determined by its similar manifestation of offenses.

The classificatory assessment of analytical variables shows that the medium-low performance of socio-behavioral, socio-morphological, and socio-political parameters characterizes an urban setting where situational crime prevention is constrained by both spatial configuration and governance practices. This condition indicates that isolated measures—such as infrastructure upgrades, landscape renewal, or policing efficiency—produce limited outcomes when not aligned with integrated strategies. The convergence of findings demonstrates that insecurity arises not from a single determinant but from the accumulation of permissive spatial arrangements, fragmented institutional actions, and social processes, some of them typical of central districts.

Thus, the results allow the inference that punctual improvement actions tend to have reduced impact if not accompanied by monitoring processes, revision of urbanistic instruments, and expanded coordination among governmental sectors. From a strategic perspective, they show that the city landscape, when planned and managed by preventive principles, can constitute significant social defense potential.

Likewise, they confirm that public security, when dissociated from urban planning solutions, loses preventive and territorialized capacity. The evidence obtained conveys that the articulation between environmental variables and institutional practices is not only desirable but essential for medium-sized cities like Itumbiara to achieve greater spatial resilience and sustainable reduction in criminal opportunities.

In addressing the proposed problem, the set of general guidelines associates analytical variables with criminological theories—both classical and contemporary. As a corollary, it confirms the hypothesis that interactions between criminal and landscape parameters enable principles for coordinated measures in planning and security in urbanized areas for the integrated management of public safety and the urban environment.

From a scientific standpoint, the discussion advances contributions related to both theoretical reinforcement and empirical demonstration of the interaction between spatial and behavioral variables, with criminality explained not only by the presence of offenders and victims but also by spatially constructed opportunities. On the other hand, by showing that thefts and robberies persist despite temporal variation and pandemic decline, the findings confirm that the landscape functions as a mediator of criminal action, allowing vulnerabilities to remain regardless of contextual dynamics.

Equally significant is the consolidation of a replicable methodological model, as the integrated reading of socio-behavioral, socio-morphological, and socio-political variables produces an original process, which, with suitable refinements, is applicable to other cities. Naturally, it requires adjustments in subsequent studies, whether within the same city or across different urban contexts. Nonetheless, preserving the simplicity of the procedures adopted is desirable, ensuring their straightforward replication not only by scholars in the field but, most notably in the Brazilian scenario, by public administrators in municipalities with limited experience in integrated processes, as well as by social actors unacquainted with specialized techniques.