Model for Measuring Urban Development with a Socioeconomic Focus in Lima, Medellin and San Salvador

Abstract

The measurement of urban development in Latin American cities has historically relied on one-dimensional models focused on GDP or density, which are insufficient to guide policies in heterogeneous and unequal contexts. This study designs and applies an empirical and comparative model with a socioeconomic approach to evaluate urban development in Lima, Medellín, and San Salvador in 2020–2024. Seventeen indicators that are comparable, objective, and publicly available in the three cities are integrated and normalized using min–max, and composite indices are estimated using three methods: equal weighting, CRITIC, and Entropy. It was found that Lima consistently shows the greatest relative progress (≈6–7%); Medellín consolidates high levels with moderate increases (≈2–4%); and San Salvador records the highest proportional growth under the Entropy method (≈11%). These findings are consistent with citizen perceptions obtained from a survey of 1170 participants (April–August 2025), according to which, on average, 60% recognize the existence of urban growth in the three cities. The proposed model is reproducible and transparent, contributing significantly to year-on-year monitoring and the prioritization of interventions by local and national governments by enabling short- and medium-term readings, in line with the Sustainable Development Goals (SDGs).

1. Introduction

The 21st century has marked a turning point in how cities have transformed the way they organize, produce, interact, and develop, becoming undisputed engines of economic growth, centers of technological innovation, and melting pots of cultural diversity. According to United Nations projections, by 2050, more than 68% of the world’s population is expected to live in urban areas [1]. This reality intensifies the interest in measuring “urban development” by incorporating socioeconomic indicators that facilitate an analysis of the territory, access to opportunities, and quality of service, among others [2,3]. This growth has generated various problems and new challenges, including socioeconomic inequality, pressure on natural resources, environmental degradation, and infrastructure congestion, among others [4,5,6]. Lima (Peru), Medellín (Colombia), and San Salvador (El Salvador) share structural challenges, such as informal land use and housing, a lack of services, and unequal mobility, among others. Each city has its own characteristics, trajectories, and scales; however, they share the same common challenge of managing urban development in an inclusive and sustainable manner.

However, most studies on urban development in Latin America have focused on “objective” indicators (socioeconomic, demographic, or infrastructural) or on descriptive qualitative analyses, with few exercises that systematically integrate citizen perceptions into comparative quantitative models. This gap limits our understanding of the extent to which urban performance as measured by official data corresponds (or does not correspond) to the way in which the population evaluates the development of their city. In this context, this study proposes a prototype model for measuring sustainable urban development for Lima, Medellín, and San Salvador (2020–2024) that combines socioeconomic indicators with citizen perception data, with the aim of articulating both dimensions in a composite index and offering a more comprehensive reading of the urban trajectories of these three Latin American cities.

Urban growth in these cities showed that looking only at traditional indicators such as gross domestic product (GDP), population density, or vehicle fleet growth rates is not enough, as they offer a partial view and, in many cases, can be misleading about what is really happening in a city [7]. Reliance on these one-dimensional indicators leads to incomplete diagnoses and, consequently, to public policies that fail to address the root causes of urban problems [8].

That is why, taking the above into account, this study addresses the central problem of the absence of a comprehensive and contextualized measurement model that allows for the evaluation of urban development from a socioeconomic perspective (covering social, economic, and environmental aspects) in the cities of Lima, Medellín, and San Salvador [9]. Without a robust and multidimensional system of indicators, decision-makers, urban planners, civil society, and academia lack a solid empirical basis for making accurate diagnoses of the strengths and weaknesses of their cities; designing evidence-based public policies that respond to the real needs of the population; monitoring the impact of interventions over time; and facilitate comparative learning and the transfer of good practices between cities with similar challenges. The guiding question of this study is: how can we build and apply a socioeconomic measurement model that allows for rigorous and comparable evaluation of the state and evolution of urban development in Lima, Medellín, and San Salvador?

To address this gap, this study proposes creating and applying a prototype model for measuring sustainable urban development in Lima, Medellín, and San Salvador for the period 2020–2024, combining objective socioeconomic indicators with citizen perception data collected through a comparative survey. Finally, at the practical level, the results of this research provide a valuable empirical diagnosis for public policy makers, civil society organizations, and citizens of Lima, Medellín, and San Salvador, fostering an informed debate on the priorities and future of development in their cities. In particular, the central methodological innovation of the study lies in the operational integration of both sources of information (socioeconomic indicators and citizen perceptions) into a composite index, which allows for a comparison between the “objective picture” of urban performance and the way residents evaluate the development of their cities.

2. State of the Art

The academic literature on measuring urban development is vast and has evolved through several stages. Initially, during the second half of the 20th century, studies were dominated by an economic paradigm, which established that a city’s development was synonymous with its economic performance. Indicators such as gross metropolitan product, foreign direct investment, and labor productivity were the central metrics [10]. While these approaches yield optimistic results in terms of a city’s economic development, they have also been strongly criticized for their “blindness” to the social and environmental consequences of urban growth [11,12].

A second wave of research was influenced by the pioneering work of the United Nations Development Program (UNDP) with the Human Development Index (HDI), which promoted multidimensional approaches, incorporating dimensions such as education (literacy rates, years of schooling) and health (life expectancy, infant mortality). This approach was adopted at the urban level, leading to the creation of human development indices at the subnational and metropolitan levels, which offered a more complete picture of the well-being of the population [13,14]. Other studies have extensively explored the adaptation of these indices to capture the specificities of urban contexts, including variables such as access to basic services such as drinking water and sanitation [15].

The third and most recent stage is guided by the idea of sustainability and resilience set out in the 2030 Agenda and its Sustainable Development Goals (SDGs), particularly SDG 11 on inclusive, safe, resilient, and sustainable cities [16]. This global framework has prompted the creation of multiple indicator systems and monitoring frameworks to assess urban growth across multiple interconnected dimensions, including economic, social, environmental, and governance [15,17], energy efficiency, waste management, quality of public space, and citizen participation [18,19,20,21,22].

In Latin America, research has followed these global trends, with a particular emphasis on addressing the structural challenges of the region. Academics have highlighted the need to incorporate indicators that capture high inequality (measured by the Gini coefficient), the extent of informal employment and housing, levels of violence, and the quality of democratic governance [23,24,25]. Studies are also being conducted on how socio-spatial segregation conditions, access to opportunities and services, and indicators are being proposed to measure the distribution of social facilities and infrastructure across the territory [26,27,28,29].

The analysis of city growth using the urban measurement models compiled in the document reveals an ecosystem of complementary frameworks that approach the city from different angles, but with the same purpose of translating urban complexity into actionable indicators.

Table 1. Existing urban development measurement frameworks.

Source	Framework	Indicators and Dimensions
[30]	UN-Habitat City Prosperity Initiative (CPI)	Comprising six dimensions, it serves to define goals and objectives that can support evidence-based policy-making, including the definition of ambitious and measurable urban visions and long-term plans.
[31]	Inter-American Development Bank (IDB) Emerging and Sustainable Cities Initiative (ESCI)	ESC employs a multidisciplinary approach to identify, organize and prioritize urban interventions to tackle the main roadblocks that prevent the sustainable growth of emerging cities in Latin America and the Caribbean. This transversal approach is based on three pillars: (i) environmental and climate change sustainability; (ii) urban sustainability; (iii) fiscal sustainability and governance.
[32]	Sustainable Development Goals (SDGs)—Localizing SDG 11. United Nations Development Programme (UNDP), UN-Habitat, local governments.	Catalogue of 400 indicators. Its purpose is to serve as a fundamental resource with the latest data and evidence for developing recommendations and solutions that enable progress toward the Sustainable Development Goals. This annual report on the SDGs is produced by the United Nations Department of Economic and Social Affairs, in collaboration with the entire United Nations Statistics System, comprising more than 50 international and regional agencies, based on data from more than 200 countries and territories.
[33]	Cómo Vamos Cities Network (Colombia and other countries)	The quality of life in cities is assessed using technical indicators, perception indicators, and monitoring of public management outcomes. It also involves the participation of experts, administrators, academics, social and community organizations, and citizens, among other development stakeholders.
[34]	Latin American Cities Climate Change Report	CCFLA offers latest research, practical resources, and tools that draw from our own expertise in the field of urban climate finance, as well as that of our members.
[35]	Sustainable Cities Index (SCI)–Mexico	It serves as a benchmark for local governments in their planning efforts. It also encourages citizens to engage with the city’s development through forums for debate on issues of collective interest, including topics such as the transformations and challenges facing the city.
[36]	Urban Resilience Index—Arup/Rockefeller Foundation. CITY RESILIENCE INDEX	Its primary purpose is to diagnose strengths and weaknesses and measure relative performance over time. This provides a holistic articulation of city resilience, structured around four dimensions, 12 goals and 52 indicators that are critical for the resilience of our cities.

shows models that reveal a transition from sectoral approaches to comprehensive, multidimensional, and multiscale urban assessment frameworks that serve both to diagnose gaps and to align local planning with global and/or state agendas.

Finally, advances in technology have opened new frontiers for analyzing and understanding the growth of cities. The field of “urban science” works with large volumes of data from different sources such as mobile phones, social networks, remote sensors, and digital transactions to construct dynamic, high-resolution indicators of mobility, economic activity, and well-being [37,38]. This process is complemented using Artificial Intelligence (AI) to analyze these data sources and extract useful patterns for urban management, as a complement to traditional statistics based on censuses and surveys [39].

Although the literature offers numerous indices and indicator systems for evaluating urban development and sustainability, there are still few proposals that articulate objective socioeconomic indicators and citizen perception data within the same analytical framework. This gap is particularly evident in the case of medium-sized Latin American cities such as Lima, Medellín, and San Salvador, where descriptive studies or studies focused on a single source of information predominate, limiting an integrated understanding of urban development processes.

3. Methodology

This study is part of a non-experimental research project with a cross-sectional design and a quantitative-comparative approach. In addition, the composite index was constructed in accordance with methodological recommendations for synthetic indicators, paying particular attention to three critical decisions: the method of normalizing the indicators, the weighting scheme adopted, and the aggregation rule. These decisions are not neutral, as they can alter the relative ranking of cities. Therefore, a parsimonious and transparent design was chosen, accompanied by a sensitivity analysis between weighting methods [40]. The methodological process includes several stages, as detailed below:

• Obtaining national statistics from each country in the socioeconomic field, with the aim of analyzing and comparing citizens’ perceptions of sustainable urban development.
• Review of indicators produced by Latin American entities such as ECLAC (UN), INE of Chile, CAF (Dominican Republic), among others.
• Construction of a prototype for measuring the degree of socioeconomic development based on socioeconomic indicators for the period 2020–2024.
• Establishing a degree of socioeconomic development for the cities of Lima, Medellín, and San Salvador.

3.1. Public Perception

The first phase of this research process consisted of developing a questionnaire comprising six questions aimed at exploring aspects related to sustainability and the level of urban development in Latin American cities. This instrument was administered simultaneously to a total sample of 1170 adults (all over the age of 18), distributed equally among three cities: Medellín (Colombia), Lima (Peru), and San Salvador (El Salvador), with 390 surveys in each. The main objective of this stage was to identify the average citizen’s perception of the degree of development of their urban environment.

From the outset, researchers recognized the existence of marked cultural and socioeconomic differences between the contexts of the selected cities, which shared only the language in general terms, albeit with notable dialectal variations. Nevertheless, it was considered feasible to identify tangential results between certain indicators, which allowed for the formulation of the study’s central hypothesis. Once the data collection process was completed in the three cities, an overview of public opinion regarding local urban development was obtained.

The survey administered in the three cities is non-probabilistic in nature and therefore the information collected includes very specific biases, which are more of a convenience type: university students, respondents from different degree programs, different occupations and personal aspirations, adults, private company office staff, middle-class families, industrial factory workers, residents of the capital from different cities, people connected to the city for various reasons, local residents, various social strata, family members, and friends of the researchers.

With the quality of responses in mind, the survey was administered to these social groups, who were considered to have a degree of reasoning and maturity. The researchers were always interested in the opinions of people with a certain level of education and, as already mentioned, preferably adults aged 18 and over. Therefore, there may be a very small sampling error, and some responses may not reflect the perception of the entire population in the three cities. However, the findings generally confirm the researchers’ personal views on the matter.

Given these limitations, the analysis of the perception survey is deliberately restricted to descriptive statistics (frequencies and percentages), without sampling weights, confidence intervals or hypothesis tests. The perception data are used in a complementary, triangulatory role, providing evidence on how citizens experience and evaluate urban development that can be compared with the patterns revealed by the composite index based on objective indicators.

3.2. Socioeconomic Indicators

The UN Millennium Development Goals can be considered one of the essential foundations of any urban development measurement model, particularly those directly related to sustainability. The nine SDGs (3–9, 11, 15) included are therefore the first reference criteria for determining which ones will be included in this study. Social and economic difficulties in most Latin American cities continue to reach critical levels despite the efforts of governments and private enterprise to resolve them. However, this study is particularly interested in identifying the problems common to the three cities under evaluation, some of which are water quality, electrical infrastructure, and vehicular traffic, among others. These characteristics are a second filter for determining the indicators that, in one way or another, are linked to sustainability and urban development, and a period of five years (2020–2024) is established to structure the model.

Therefore, the initially proposed indicators are directly related to the UN Millennium Development Goals and sustainability. They must measure the same variables in the three cities using uniform units and also be available on the internet (public access). The selection of indicators is also informed by the SIEDU system documented by INE (2025) [41], which analyses 90 indicators covering environmental, social, economic, urban development, education and health dimensions. Within this framework, 26 indicators are classified as urbanistic, 4 as architectural and heritage-related, and 60 as socioeconomic. The 17 indicators used in this study belong to this last group and were chosen because they are consistently available, comparable across the three metropolitan areas and both years, and directly related to the conceptual dimensions of our urban development framework.

The 17 indicators are constructed from official statistics and administrative records produced by national and local institutions in each country (

Table 2. Model for measuring socioeconomic development 2020–2024 with indicators grouped by type of economic development.

Indicators	Medellín		Lima		San Salvador
	2020	2024	2020	2024	2020	2024
Economic Development
PIB per capita relative to PIB in developed countries	17.60%	13.80%	13.00%	23.00%	10.00%	11.00%
2. Female labor force participation	56.10%	57.60%	55.00%	63.00%	47.00%	49.00%
Social Welfare and Cohesion
3. Families with their own homes	91.90%	96.20%	76.00%	76.00%	54.00%	58.00%
4. Employment rate for women and men of all ages	78.40%	93.50%	64.00%	70.00%	95.00%	95.00%
5. Average schooling of 12 years in total	89.20%	90.80%	69.00%	69.00%	69.00%	67.00%
6. Population in the AMSS with access to health insurance	97.80%	98.60%	77.00%	88.00%	27.00%	29.00%
7. School enrollment	92.00%	94.00%	96.00%	98.80%	19.00%	16.00%
8. Child survival rate for children under 5	99.72%	99.20%	98.00%	98.00%	99.00%	99.00%
Urban Sustainability and Infrastructure
9. Number of households with Internet access	56.50%	60.70%	40.00%	58.00%	57.00%	46.00%
10. Vehicle fleet	54.80%	60.00%	21.00%	21.00%	20.00%	29.00%
11. Square meters of green space per capita (UN 15 m2 = 100%)	26.70%	26.70%	17.10%	17.10%	22.00%	22.00%
12. SDG 6: Clean Water and Sanitation	95.30%	96.10%	91.00%	90.00%	79.00%	96.00%
13. SDG 7: Affordable and Clean Energy	97.00%	99.00%	96.00%	97.00%	83.00%	80.00%
Governance and Institutional Capacity
14. National Budget for Education	16.20%	14.10%	4.26%	4.60%	4.17%	4.43%
15. National Budget for Health	13.40%	13.30%	4.35%	3.99%	3.04%	3.40%
16. Researchers (innovation/knowledge)	0.289%	0.28%	0.01%	0.03%	0.30%	0.16%
17. Metropolitan Area Population in relation to the total population of the country	8.01%	7.89%	29.65%	30.61%	28.82%	37.93%

). For Medellín, the main sources are DANE and municipal reports on housing, labour markets, social protection and basic services. For Lima, we rely on the INEI population census (2017) and household surveys, national accounts and sectoral budget data, complemented by CONCYTEC and ministerial statistics. For San Salvador, we use the BCR–DIGESTYC Multipurpose Household Survey (EHPM), sectoral reports compiled by ICEFI, ORMUSA and PAHO/WHO, the national vehicle registry and fiscal transparency reports.

While the composite index provides objective, quantifiable, and comparable measures that are useful for public and private decision-making, an inherent limitation of this type of model is its difficulty in capturing the human experience of urban space. The indicators quantify levels of provision (e.g., square meters of green space per capita), but not necessarily the perceived quality or condition of those spaces. For example, a city may meet the United Nations recommendation of 9 m² of green space per person, but those areas can range from well-maintained parks to underutilized or degraded vacant lots. This motivates the incorporation of perception data into our framework and suggests that future extensions should further explore how subjective assessments can complement objective measures of urban development.

The authors are aware of the limitations of a model, but in other words, a model is the equivalent of a technical and strategic diagnosis that includes comparative analyses between cities to detect what could be some keys to stimulating urban development in certain communities, and this is one of the objectives of constructing this model.

3.3. Methods of Measuring Urban Development

Three methods were chosen for weighting and calculating the composite index: equal weighting, CRITIC, and Entropy. The literature on composite indicators warns that weight allocation is one of the most controversial decisions, as it can introduce strong normative judgments or, at the opposite extreme, completely delegate the importance of each indicator to its statistical properties [40]. Therefore, a “neutral” reference scheme (equal weighting) was combined with two objective weighting methods (CRITIC and Entropy), which allows the robustness of the results to be evaluated against alternative assumptions about the relative importance of the variables. The CRITIC and Entropy methods belong to objective weighting schemes, whereby the weights of each indicator are derived directly from the statistical properties of the data. CRITIC exploits internal contrast (variability) and conflict between criteria, while Entropy captures the amount of effective information in each indicator. In this way, both methods assign greater weight to those variables that truly differentiate the performance of cities and reduce redundancy between indicators.

In operational terms, the three methods share the same standardized data matrix, but differ in how they extract the relative importance of each indicator. This makes it possible to distinguish between changes in the index attributable to “real” variations in the data and changes due exclusively to the weighting scheme adopted. This strategy responds to the recommendation to accompany the construction of composite indices with sensitivity exercises that explore different sets of weights. Each method is described below:

3.3.1. Equal Weighting Method

This method assigns a uniform weight to each alternative or indicator, without considering their variability or internal correlation [40]. Let $x_{ij}$ denote the normalized value (between 0 and 1) of indicator $j=1,\dots , K$ for city $i=1,\dots , N$ in a given year. With equal weighting, each indicator receives the same weight $w_j=1/K$. The composite urban development index for city $i$ is then

$$I_i^{EW}=\sum _{j=1}^K{w}_i,x_{ij}={\displaystyle \frac{1}{K}}\sum _{j=1}^K{x}_{ij}$$

(1)

This scheme provides a transparent baseline in which all indicators contribute equally to the index, consistent with the conceptual framework that treats the four dimensions of development as equally important.

3.3.2. Criteria Importance Through Intercriteria Correlation (CRITIC)

Proposed by Diakoulaki et al. [42], this method aims to calculate the weights of multi-criteria problems, combining two main ideas. The first relates to the internal contrast of each criterion (variability between alternatives). The second relates to the conflict with the other criteria. In this way, the criteria that differentiate the most and repeat the least information predominate.

Formally, consider the decision matrix $X=\left(x_{ij}\right)$ with $N$ cities and $K$ indicators, where $x_{ij}$ are the min–max normalized values. Let ${\sigma }_j$ be the standard deviation of indicator $j$ across cities and years, and let $r_{jk}$ be the Pearson correlation coefficient between indicators $j$ and $k$. The amount of information or contrast associated with indicator $j$ is computed as

$$C_j={\sigma }_j\sum _{k=1}^K(1-r_{jk}).$$

(2)

Indicators that are more dispersed (${\sigma }_j large)$ and less correlated with the rest (low $r_{jk}$) receive higher $C_j$ values, reflecting greater discriminatory power and lower redundancy. The CRITIC weights are then obtained by normalizing these information measures:

$$w_j^{CRITIC}={\displaystyle \frac{C_j}{\sum _{m=1}^K\left(c_m\right)}}, j=1,\dots , K.$$

(3)

Finally, the composite urban development index for city $i$ under CRITIC weighting is

$$I_i^{CRITIC}=\sum _{j=1}^K\left(w_j^{CRITIC}{x}_{ij}\right).$$

(4)

In this way, CRITIC emphasizes indicators that most clearly distinguish between cities and years while penalizing redundant information, which is particularly useful for highlighting effective performance gaps in comparative analysis.

3.3.3. Entropy

The entropy method is based on Shannon’s notion of entropy [43] as a measure of uncertainty. This method captures how much indeterminacy there is in the data. On that basis, entropy weighting was introduced into the multicriteria decision-making literature as an “objective” scheme that assigns greater weight to indicators that show greater differentiation between alternatives. Following the standard Shannon entropy implementation in multi-criteria decision-making, the first step is to convert the normalized indicators into proportions:

$$p_{ij}= {\displaystyle \frac{x_{ij}}{\sum _{i=1}^N{x}_{ij}}}, i=1,\dots , N; j=1,\dots ,K.$$

(5)

The entropy of indicator $j$ is then

$$e_j=-k\sum _{i=1}^N{p}_{ij}\mathit{ln}\left(p_{ij}\right), k={\displaystyle \frac{1}{\mathit{ln}N}},$$

(6)

where $k$ is a normalization constant ensuring that $0\le e_j\le 1$. Indicators that are more evenly distributed across cities (low dispersion) have higher entropy, whereas indicators with more unequal distributions (high dispersion) have lower entropy and therefore carry more information. The degree of divergence (or information) of indicator $j$ is

$$d_j=1-e_j.$$

(7)

The entropy weights are obtained by normalizing these divergence measures:

$$w_j^{ENT}={\displaystyle \frac{d_j}{\sum _{m=1}^K\left(d_m\right)}}, j=1,\dots , K.$$

(8)

As in the other schemes, the composite index for city $i$ is

$$I_i^{ENT}=\sum _{j=1}^K\left(w_j^{ENT}{x}_{ij}\right).$$

(9)

This method assigns higher weights to indicators that display greater relative heterogeneity across cities and years, highlighting structural capacities and underlying imbalances in the urban development process.

Although there are other very popular methods for this context, such as Principal Component Analysis (PCA) [44] or schemes based on expert opinion (e.g., AHP or Delphi methods) [45,46], these have not been included for one or more reasons. For example, PCA requires a sufficient number of observations to estimate stable components; in our case, the city-year matrix consists of only six alternatives (three cities in two years), which could produce solutions that are not very robust and difficult to interpret in terms of urban policy and the goals associated with the SDGs. For their part, expert-based methods require the formation of representative panels in each country, which introduces potential power asymmetries between cities and limits the replicability of the model by other institutions. It is not ruled out that future work will explore hybrid schemes that combine objective weightings with expert judgments; however, the objective of this prototype was to prioritize a parsimonious design that is replicable and easily auditable by third parties.

4. Results

This section presents the results obtained after applying the data collection instruments, which were processed using SPSS statistical software (V.26). The urban development measurement methods were implemented in the Python v.3.12.7 programming language and executed under a 64-bit Windows 11 operating system, with an Intel Core I9 processor, 32GM RAM, and a GEFORCE RTX 5080 video card.

4.1. Analysis of Perceptions of Urban Development and Sustainability

The survey conducted in Lima, Medellín, and San Salvador provides an empirical exploration of citizens’ attitudes toward different aspects of sustainable urban development. The main conclusions reached are as follows:

• An average of 60% of respondents in the three cities agree that there is a certain degree of urban growth.
• Forty-three percent of the total sample think that urban development is not going in the direction they would really like.
• Only 42% of the total sample think that concrete efforts (urban actions or projects) are being made in their city to achieve sustainable development.
• 49% of the average citizen thinks that their city can become sustainable someday in the future.
• 60% of respondents say that public space in cities is not distributed equitably.
• 61% of the sample believes that the benefits of development are not being distributed equitably in their cities.

These results reveal a positive outlook for sustainable urban development in these cities. Although there is still a long way to go, the attitude of their citizens encourages efforts to achieve this goal. It also alerts local governments to widespread dissatisfaction among the population due to the poor distribution of public space and resources.

4.2. Urban Development Statistics

As specified in Section 3.2, urban development measurement methods require that the data for the indicators set out in Table 2 be on a standardized scale. To this end, minimum and maximum normalization was applied, also ensuring the directionality of benefit (higher values imply better performance). This step allows for subsequent aggregation and intertemporal comparison between cities for 2020–2024 (

Table 3. Indicators normalized on a scale of 0 to 1, using the min-max method.

Indicators	Medellín		Lima		San Salvador
	2020	2024	2020	2024	2020	2024
Economic Development
1. PIB per capita relative to PIB in developed countries	0.176	0.138	0.130	0.230	0.100	0.110
2. Female labor force participation	0.561	0.576	0.550	0.630	0.470	0.490
Social Welfare and Cohesion
3. Families with their own homes	0.919	0.962	0.760	0.760	0.540	0.580
4. Employment rate for women and men of all ages	0.784	0.935	0.640	0.700	0.950	0.950
5. Average schooling of 12 years in total	0.892	0.908	0.690	0.690	0.690	0.670
6. Population in the AMSS with access to health insurance	0.978	0.986	0.770	0.880	0.270	0.290
7. School enrollment	0.920	0.940	0.960	0.988	0.190	0.160
8. Child survival rate for children under 5	0.9972	0.992	0.980	0.980	0.990	0.990
Urban Sustainability and Infrastructure
9. Number of households with Internet access	0.565	0.607	0.400	0.580	0.570	0.460
10. Vehicle fleet	0.548	0.600	0.210	0.210	0.200	0.290
11. Square meters of green space per capita (UN 15 m2 = 100%)	0.267	0.267	0.171	0.171	0.220	0.220
12. SDG 6: Clean Water and Sanitation	0.953	0.961	0.910	0.900	0.790	0.960
13. SDG 7: Affordable and Clean Energy	0.970	0.990	0.960	0.970	0.830	0.80
Governance and Institutional Capacity
14. National Budget for Education	0.162	0.141	0.0426	0.046	0.0417	0.0443
15. National Budget for Health	0.134	0.133	0.0435	0.0399	0.0304	0.034
16. Researchers (innovation/knowledge)	0.00289	0.0028	0.0001	0.0003	0.003	0.0016
17. Metropolitan Area Population in relation to the total population of the country	0.08010	0.0789	0.2965	0.3061	0.2882	0.3793

).

Sensitivity analysis allows us to refine the interpretation of priorities based on the weights obtained for each method (

Table 4. Weights per indicator obtained by urban development methods.

Indicators	Equal Weighting	CRITIC	ENTROPY
Economic Development
1. PIB per capita relative to PIB in developed countries	0.0588	0.0288	0.0314
2. Female labor force participation	0.0588	0.0315	0.0035
Social Welfare and Cohesion
3. Families with their own homes	0.0588	0.0632	0.0162
4. Employment rate for women and men of all ages	0.0588	0.1035	0.0087
5. Average schooling of 12 years in total	0.0588	0.0385	0.0065
6. Population in the AMSS with access to health insurance	0.0588	0.1394	0.0802
7. School enrollment	0.0588	0.2046	0.1298
8. Child survival rate for children under 5	0.0588	0.0036	0.0000
Urban Sustainability and Infrastructure
9. Number of households with Internet access	0.0588	0.0407	0.0076
10. Vehicle fleet	0.0588	0.0651	0.0825
11. Square meters of green space per capita (UN 15 m2 = 100%)	0.0588	0.0192	0.0119
12. SDG 6: Clean Water and Sanitation	0.0588	0.0338	0.0016
13. SDG 7: Affordable and Clean Energy	0.0588	0.0387	0.0025
Governance and Institutional Capacity
14. National Budget for Education	0.0588	0.0195	0.1408
15. National Budget for Health	0.0588	0.0172	0.1506
16. Researchers (innovation/knowledge)	0.0588	0.0008	0.2235
17. Metropolitan Area Population in relation to the total population of the country	0.0588	0.1520	0.1028

). The equal weighting baseline assigns a weight of 0.0588 to all indicators. In contrast, CRITIC concentrates on weight on variables with high contrast and low redundancy: school enrollment (0.2046), access to health insurance (0.1394), metropolitan share of the national population (0.1520), and employment rate (0.1035). At the opposite end of the spectrum, infant survival (0.0036), number of researchers (0.0008), and green areas per capita (0.0192) have marginal weight, consistent with their low relative variation between cities. Entropy shifts the weight toward “informationally rich” attributes: researchers (0.2235), health budget (0.1506), education budget (0.1408), enrollment (0.1298), metropolitan share (0.1028), vehicle fleet (0.0825) and, to a lesser extent, access to insurance (0.0802).

These results confirm that the Equal Weighting method offers a neutral benchmark for all indicators analyzed. CRITIC prioritizes effective performance gaps (educational and insurance coverage, labor market insertion, and metropolitan structure), which are useful for targeting interventions where the greatest differences exist. Finally, Entropy favors structural capacity bases (R&D, social fiscal effort, metropolitan mass), revealing medium-term levers to sustain progress. The almost zero weights in child survival and SDGs 6–7 respond to their high saturation and low interurban dispersion, while Internet and green areas retain reduced weight due to their lower discriminatory power.

To quantify urban development trends between the periods studied (

Table 5. Urban development index 2020–2024.

City	Equal Weighting			CRITIC			ENTROPY
	2020	2024	% Δ	2020	2024	% Δ	2020	2024	% Δ
Lima	0.5008	0.5342	6.6669	0.6169	0.6584	6.7278	0.2840	0.3028	6.5955
Medellín	0.5829	0.6010	3.1134	0.6756	0.7046	4.3006	0.3413	0.3470	1.6508
San Salvador	0.4220	0.4370	3.5674	0.4011	0.4204	4.8013	0.1402	0.1555	10.9596

), the percentage change $(\mathrm{\%} \mathsf{\Delta })$ in the composite indices was calculated using Equation (10).

$$\mathrm{\%}\mathrm{ }\mathsf{\Delta }=\left[{\displaystyle \frac{{Index}_{c,2024}-{Index}_{c,2020}}{{Index}_{c,2020}}}\right]\times 100$$

(10)

where ${Index}_{c,t}$ represents the value of the index for city $c$ in year $t$. It should be noted that, because these indices are constructed from aggregate macroeconomic and social indicators from official sources (deterministic and non-stochastic data for each cut-off year), the variations reported describe the magnitude of the relative change in the performance of the normalized indicators and are not subject to traditional inferential statistical significance tests derived from probabilistic sampling.

When analyzing the magnitude of these variations, different behaviors are observed by city, which respond to the nature of the methods applied. Lima shows the most consistent and homogeneous growth among the three methods ($\approx$6.6–6.7%), suggesting comprehensive and balanced progress in both basic indicators and those with greater structural variability. Medellín, which has the highest absolute scores, records more moderate growth rates (1.65% to 4.30%), a behavior typical of diminishing marginal returns where it is more difficult to achieve large percentage increases when starting from a consolidated base of development. Finally, San Salvador shows the greatest sensitivity to methodological choice. Although its growth is moderately low under Equal Weighting and CRITIC (3.6–4.8%), it experiences a significant jump under the Entropy method (10.96%). This is because Entropy assigns high weights to variables with high uncertainty or differentiation, such as “Researchers” and “Health Budget,” areas where San Salvador, starting from minimal bases, has recorded proportional changes that the model interprets as a significant gain in information and relative development.

Using the six observations presented in Table 5, we compute pairwise Pearson correlation coefficients between the three versions of the index (equal weighting, CRITIC and Entropy weighting). The resulting coefficients are $r_{E,C}=0.956$ for the correlation between the equal-weighted and CRITIC-based indices, $r_{E,T}=0.971$ between the equal-weighted and Entropy-based indices, and $r_{C,T}=0.996$ between the CRITIC- and Entropy-based indices. These very high positive correlations indicate that, within the limits imposed by the small number of analytical units, both the levels of the composite index and the relative differences between cities are robust to the particular weighting method adopted.

Taking into account the results, it can be seen that, in economic terms, Lima has the highest relative GDP per capita, allowing for advantages in the expansion of health services, digital connectivity, and access to medical insurance. Meanwhile, Medellín, which has an intermediate GDP, rewards urban development by contributing to innovation and sustained planning. In contrast, San Salvador has a lower GDP, which allows for public infrastructure coverage, quality urban services, and the state’s ability to sustain long-term social policies.

With regard to the labor market and women’s participation, Medellín shows progress thanks to the implementation of municipal programs and strategic alliances that promote female employability and entrepreneurship. Similarly, in Lima, female labor participation is significant, but with high levels of informality, which limits access to social security. In addition, San Salvador has greater gaps in employment and female participation, generating problems of security, inequality, and opportunities in the formal labor market.

In terms of social welfare and cohesion indicators, Medellín stands out for its educational and public health policies, which show relatively high levels of average schooling, school enrollment, and child survival. Similarly, Lima shows significant progress, especially in access to health insurance and educational coverage, but territorial heterogeneity conditions expected achievements. In addition, San Salvador faces greater limitations in educational continuity, determined by fiscal constraints, urban poverty, and migratory flows that stabilize its social protection systems.

In terms of urban sustainability, Medellín stands out for its green infrastructure, access to services, and fulfillment of goals associated with SDGs 6 and 7, the result of environmental policies sustained for more than two decades, led by the institutional ecosystem of Public Companies of Medellín. Lima shows progress in digital connectivity and services, but there is still a deficit in green areas per capita and in the provision of sanitation in peripheral areas. Meanwhile, San Salvador has significant gaps in water, sanitation, and urban infrastructure, which limit its ability to meet international sustainability standards.

Finally, in terms of institutional capacity, Medellín has solid government structures, efficient budget execution, and an innovation ecosystem that integrates universities, the private sector, and local government. Similarly, Lima depends to a greater extent on the central government, whose political variability affects the continuity of sectoral policies in health, education, and infrastructure. Likewise, San Salvador faces fiscal constraints and a lower availability of specialized personnel, in terms of the number of researchers, the quality of public management, and service outcomes.

5. Discussion

The urban planning concept of urban sustainability is still a utopia in most Latin American cities, which is one of the reasons why this research is justified. This prototype is a tool that could be very useful to all entities interested in measuring the degree of development in any city.

This discussion analyzes the main findings obtained in relation to the objectives set. First, a positive condition is observed in the sustainability of urban development in the three cities under study, Medellín, Lima, and San Salvador, which coincides with the idea of sustainability and resilience of the 2030 Agenda and SDG 11 on inclusive, safe, resilient, and sustainable cities [16]. However, various problems and new challenges were identified, notably socioeconomic inequality, pressure on natural resources, environmental degradation, and infrastructure congestion, among others [4,5,6].

The findings of this study have significant theoretical implications, contributing to the literature by proposing a prototype measurement model for three cities. On a practical level, it focuses on empirical diagnosis that is valuable to public policy makers, civil society organizations, and citizens of the three cities under study, enabling the prioritization and development of the cities.

Despite its contributions, the study has some limitations in terms of the population that is dissatisfied with public spaces, resources, and development benefits because they are not distributed equitably. Likewise, urban development is not going in the direction they want.

These limitations may affect the results, so future research should incorporate indicators that measure high inequality, labor and residential informality, levels of violence, and the quality of democratic governance [9,30,31]. In addition to proposing indicators that measure the distribution of social facilities and infrastructure in the territory [32,33,34,35], future research should use artificial intelligence to extract useful patterns for urban management [45].

The study shows that building a measurement model based solely on socioeconomic indicators generates a distorted view of reality and that, therefore, decision-making based on this information can lead to misguided urban or physical outcomes.

Those responsible for the development and growth of cities must not forget that an essential dimension inherent to urban planning is the spatial variable defined by infrastructure and architecture, and that the physical variables of a city are many and have diverse impacts on the environment. The main ones must be considered when constructing a comprehensive measurement model.

In this regard, both the state and private enterprise play a leading role in designing public policies and implementing them in a city. What is not seen in this measurement model, but should not be overlooked, is that these public policies must be backed by regulations that control and provide feedback on the construction and planning process in order to correct mistakes.

One of the objectives of the study is to compare development between cities, and from the outset it was assumed that the measurement model and mathematical methods applied would facilitate this task. Indeed, measuring and comparing variations in urban development in 2020 and 2024 in three different cities prompts reflection on the successes and failures of urban planning in Latin America. In summary, the results show the relative positioning of each city, indicated not only by the numerical values of the indicators, but also by structural conditions, institutional capacity, public investment, and the strategic orientation of urban policies.

6. Conclusions

The study demonstrates that an empirical model with a socioeconomic approach can be used to evaluate and compare the urban development of three cities: Lima, Medellin, and San Salvador. Initially, a questionnaire was conducted to identify the degree of urban growth according to citizens’ perceptions.

Subsequently, a model was developed including 17 indicators and evaluated with three weighting models, which offers a more detailed reading of urban development than traditional one-dimensional models. The application of the model to three Latin American cities corroborates the usefulness of combining “neutral” and “objective” methods to distinguish between short-term progress and structural capacities, while maintaining traceability and replicability for year-on-year monitoring.

The results are consistent in terms of performance ranking: Medellín remains in first place, followed by Lima and then San Salvador. This quantitative evidence is consistent with the findings on citizen perception, which recognizes growth but expresses doubts about its direction and equitable distribution. In terms of management, the proposed model offers a replicable basis for annual scorecards that allow cities to be compared, goals to be monitored, and investments to be prioritized in line with the SDGs.

However, there are limitations that open up clear avenues for future research. The number of indicators analyzed is relevant but limited; incorporating metrics of spatial segregation, quality of public space, safety, and climate change would strengthen external validity and open up the possibility of proposing a new urban measurement model that surpasses the methods evaluated here. Likewise, the integration of emerging data sources (remote sensing, mobile telephony, transactions, among others) and AI techniques for imputation and outlier detection would improve the coverage and timeliness of information through urban scenario simulation, detection of urban inequality through data analysis, optimization of smart urban services, and analysis of digital and socioeconomic gaps.