Regional Economics · Post-War Recovery · JEL O18, J23

Classification of Territorial Communities by the Level of Socio-Economic Potential

Methods and models of assessment in the context of post-war recovery — a taxonomic and cluster-analysis approach for ranking Ukraine's territorial communities and identifying the most vulnerable.

OROlha Rudachenko VSValentyna Smachylo OKOleh Kulinich DZDenys Zakharov
JEL O18 JEL J23 territorial communities socio-economic potential sustainable development post-war reconstruction human capital investment entrepreneurship
Total in Ukraine
1469
territorial communities
Pareto sample
300
communities studied (≈20%)
Cluster analysis
8
clusters identified
Classification
3
potential groups — A, B, V

Abstract

A methodology for ranking communities and finding those most at risk

Post-war recovery of territorial communities (TC) is a key task for ensuring the country's stable development. The war has significantly impacted the socio-economic condition of communities, causing substantial losses in infrastructure, human capital, and economic activity.

In these circumstances, there is a need for a practical assessment of community potential to determine support priorities and develop development strategies. This article is dedicated to creating a methodological approach for classifying territorial communities based on their socio-economic potential (SEP), which serves as a basis for ranking and identifying the most vulnerable communities in the context of post-war recovery. The authors analyze community development factors and propose an evaluation system that considers infrastructure, demographic, economic, and social aspects.

The methodological approach is based on taxonomic analysis and clustering, which allows for identifying groups of communities with similar characteristics and determining priority development directions. The study examines key evaluation criteria, including the level of economic activity, accessibility of essential services, employment levels, and the quality of social infrastructure. The proposed methodology classifies communities by their development level and forms targeted support strategies tailored to the specific needs of areas. This ensures the effective use of state and international aid resources, promotes investment attraction, and fosters the formation of sustainable economic recovery models.

Rank, don't guess

Integral SEP indicators let scarce state and international aid be directed to the communities that need it most.

Group by similarity

Clustering reveals data structures invisible to the eye, grouping communities into tiers of comparable potential.

Tailor the strategy

Each tier receives recovery measures matched to its specific needs — from infrastructure to human capital.

Problem statement & aim

No single methodology yet exists for communities

Effective rehabilitation and development require a comprehensive assessment of socio-economic potential. Classifying communities by potential level lets resources and support concentrate where they are most needed.

A community's socio-economic potential defines its capacity for recovery and development. Classification clarifies which communities need priority assistance for restoring infrastructure, developing entrepreneurship and human capital, and improving social conditions. Such assessment is also essential for planning state financial policy at the local level, allocating resources, and stimulating economic growth.

One of the main problems in this field today is the absence of a single, effective methodology usable across territorial communities, which complicates decision-making on recovery and development both during and after the war. Existing assessment models often fail to account for a community's specifics — damaged infrastructure, lost human capital, and changed social, ecological, and economic conditions. There is also a shortage of studies that integrate the different aspects of local socio-economic development and the specific challenges faced by affected communities.

The gap this study addresses

New approaches are needed that not only assess the general level of socio-economic potential, but also account for the particular conditions of post-war recovery — including the need to adapt economic and social structures to new realities.

Aim of the article

To form a methodological approach for classifying territorial communities by the level of their socio-economic potential — serving as a basis for ranking and identifying the most affected communities.

Research tasks

Analyze the general state of communities during wartime; develop a classification methodology that accounts for post-war recovery; and propose practical recommendations for planning, development strategy, and recovery plans.

Methodological approach · Fig. 1

Six stages, from selection to recommendations

Classification by socio-economic potential aims to reveal data structures by forming groups of similar objects — clusters — making patterns visible that observation alone would miss. The proposed approach proceeds through six stages.

1

Selection of communities

Select the territorial communities for study.

2

Forming indicators

Build the set of indicators that characterize the SEP of communities.

3

Integral indicators

Construct integral SEP indicators using taxonomy methods.

4

Classification by SEP

Classify communities by their integral level of socio-economic potential.

5

Distribution by class

Distribute the communities across their corresponding classes.

6

Recommendations

Formulate recommendations for the formed SEP classes of communities.

Stage 1 · The Pareto sample · Fig. 2

Studying 20% to understand the whole

Of 1469 territorial communities in Ukraine, the authors randomly selected 300 whose statistical data are fully available in official open sources. This follows the principle of V. Pareto and R. Koch's "80/20" rule — that 80% of results stem from 20% of efforts or causes — widely applied in education and science, business, time management, personal effectiveness, sustainable development, and territorial governance.

Implementation of the Pareto principle in selecting communitiesFigure 2

Source: compiled by the authors.

Stage 2 · Indicators · Fig. 3

Two blocks: economic and social

To assess SEP comprehensively, the authors analyzed numerous international indices — the happiness index, the gender inequality index, the human development index, and others — then grouped indicators into economic and social blocks reflecting key aspects of community life.

Economic block

  • Tax revenuesUAH
  • Number of business entitiesunits
  • Average monthly wageUAH
  • InvestmentsUAH
  • TransfersUAH
  • Receipts under aid programs (EU, foreign governments, international & donor organizations)UAH
  • Consumer price index%
  • Community incomeUAH

Social block

  • Areakm²
  • Number of unemployed registered at period endpersons
  • Working-age populationpersons
  • Community populationpersons

This system of indicators is not always informationally accessible, reliable, regularly updated, or expandable over time and space — which motivates the construction of integral indicators in the next stage.

Stages 3–4 · Taxonomy & cluster analysis

From raw indicators to integral scores and clusters

Using taxonomy methods, integral SEP indicators were built for the communities. Analysis of these values showed a negative trend and an annual decline in socio-economic potential.

In most of the studied communities, indicators lag significantly behind the maximum value of 1, which serves as the benchmark in the proposed models — the closer the integral indicator is to 1, the higher the level of socio-economic potential. The chart below compares the integral SEP indicators of communities in the Mykolaiv region for 2022–2023.

Integral SEP indicators — Mykolaiv region communities, 2022 vs 2023Figure 4

Source: compiled by the authors.

Distance measures & the k-means objective

Classification by the integral SEP indicator used cluster-merging techniques and algorithms. The Euclidean distance between two points x and y is the shortest distance between them; in two- or three-dimensional space it is the geometric length of the line segment connecting the points. For n variables it is computed by formula (1):

$$ dist(x,y)=\sqrt{\sum_{i=1}^{n}(x_i - y_i)} $$(1)

The square of the Euclidean distance is defined by formula (2):

$$ dist(x,y)=\sum_{i=1}^{n}(x_i - y_i)^2 $$(2)

Hierarchical clustering was performed by the furthest-neighbour (complete-linkage) method using the squared Euclidean distance. In such a dendrogram, the merge height shows the degree of similarity between objects: the lower the merge height, the more similar the objects. This allows the number of clusters to be determined visually — a clear split into 8 clusters was observed.

A two-step clustering algorithm was then applied — designed for large data volumes and working effectively with both continuous and categorical variables. Among iterative methods, the k-means algorithm (fast cluster analysis) requires the probable number of clusters to be set in advance. The method minimizes the sum of squared distances between each observation and its cluster center — a function that evaluates clustering quality, formulas (3)–(4):

$$ \sum_{i=1}^{k}\;\sum_{k\in D_i}\lVert x_i - c_i\rVert^{2} $$(3)

where Di is the set of vectors belonging to cluster i, and ci is the mean value of these vectors.

$$ c_i=\frac{\sum_{k=1}^{N_i} x_k}{N_i},\qquad x_k\in D_i $$(4)

On each iteration a new center of mass is computed for every cluster; data are then reassigned to the nearest center under the chosen metric. The process repeats until changes become negligible or the centers stop moving.

Results · Tables 1–3, Fig. 6

Eight clusters, grouped into three tiers

Based on the obtained integral indicators, 8 clusters were formed and combined into 3 groups. Cluster centers shifted from their initial to final positions across the k-means iterations.

Cluster centers of the SEP integral index — initial vs finalTables 1 & 2

Source: compiled by the authors. Bars are coloured by tier — Group A (clusters 1–3), Group B (clusters 4–6), Group V (clusters 7–8).

189.04
F-statistic
0.000
Significance
7 / 123
Degrees of freedom (cluster / error)
0.123
Cluster mean square (error 0.001)

Table 3 — Analysis of variance. The F-statistic should be read only as an indicator: clusters were chosen to maximize differences between observations, so the observed significance levels were not adjusted and cannot be used to test the hypothesis of equal cluster means.

The three potential groups

AHigh socio-economic potentialClusters 1 · 2 · 3

Developed infrastructure, an active business sector, and low unemployment. These communities receive significant funding from the state and international organizations, making them socio-economically stable and prosperous.

Cluster 10.686
Cluster 20.490
Cluster 30.454
BMedium socio-economic potentialClusters 4 · 5 · 6

Limited investment, but transfers partially support the economy. These communities have potential for development — provided there is proper support and financing.

Cluster 40.255
Cluster 50.202
Cluster 60.134
VLow socio-economic potentialClusters 7 · 8

High unemployment, low incomes, a shortage of enterprises and financing. A crisis situation with limited investment and transfers — requiring deeper research and targeted measures to resolve socio-economic problems.

Cluster 70.101
Cluster 80.090
Why Group V matters most

Communities in Group V typically face limited access to financial resources, which constrains infrastructure projects and quality social services. Low investment signals weak business attractiveness, limiting new jobs — and low potential drives population outflow, especially of youth seeking better conditions elsewhere. These communities need targeted support to raise their stability and viability, particularly under wartime and post-war recovery.

Distribution of communities by cluster

The matrix below distributes the studied territorial communities (2023) across the eight clusters and three groups.

A Group A — high potential
Cluster 1
Dniprovska
Cluster 2
Odeska
Cluster 3
KryvorizkaMykolaivska
B Group B — medium potential
Cluster 4
Mariupolska
Cluster 5
BerdianskaMelitopolskaBorshchahivskaLysychanska
Cluster 6
IvankivskaKakhovska
V Group V — low potential
Cluster 7 · 19 communities
MahdalynivskaMarhanetskaNikopolskaKostiantynivskaNizhynskaBashtanskaIzmailskaShabivskaKonotopskaOkhtyrskaTrostianetskaBalakliiskaRohanskaVysochanskaIziumskaValkivskaKrasnohradskaBilozerskaBakhmatska
Cluster 8 — the largest cluster
SlobozhanskaVasylkivskaKarpivskaLiubymivskaNovomoskovskaDruzhkivskaStarobilskaMyrnohradskaNikolskaOlhynskaPokrovskaSartanskaSviatohirskaSlovianskaAndriivskaVasylivskaVeselivskaKyrylivskaKushuhumskaMalynivskaNovobohdanivskaOrikhivskaBerezanskaBoryspilskaBoiarskaVyshnevaHirskaMakarivskaObukhivskaPoliskaSkvyrskaTashanskaKreminskaMilovskaPopasnianskaSvativskaSievierodonetskaTroitskaShchastynskaArbuzynskaVesnianskaVoznesenskaHorokhivskaNovobuzkaOchakivskaPrybuzkaSofiivskaStepivskaYuzhnoukrainskaLymanskaMykolaivskaOvidiopolskaPavlivskaPodilskaReniiskaSavranskaSaratskaSerhiivskaStarokozatskaTeplytskaTuzlivskaBerezivskaBoromlianskaVilshanskaHlukhivskaDruzhbivskaLebedynskaRomenskaMerefianskaStarovirivskaNovovodolazkaZmiivskaDvorichanskaKehychivskaPervomaiskaSakhnovshchynskaAskania-NovaVynohradivskaHenicheskaHoloprystanskaDolmativskaLazurnenskaMylivskaOleshkivskaBaturynskaBereznianskaBorznianskaHoncharivskaDesnianskaDobrianskaKoropskaLadanskaLynovytskaMakiivskaOsterskaRipkynskaYablunivska

Conclusions · for Group V communities

Seven measures to lift the most vulnerable

The proposed methodology identified communities of high (Group A), medium (Group B), and low (Group V) potential. Group V — clusters 7 and 8 — needs comprehensive measures to overcome the crisis and create conditions for stable development.

Economic recovery strategies

Create long-term development strategies focused on attracting investment, supporting local business, and creating jobs — including programs for entrepreneurship and startups to gradually restore economic activity.

Additional financial resources

Actively seek extra funding through state and international grants, international programs, and other support mechanisms to cover gaps in social and infrastructure projects.

Human capital development

Organize retraining and upskilling programs for the unemployed and youth, improving their labour-market chances and reducing unemployment.

Investment attractiveness

Build favourable conditions for investors: simplify administrative procedures, lower tax rates for small and medium business, and develop investor-friendly infrastructure.

Social programs

Improve social infrastructure — access to quality medical services, education, and social protection for the most vulnerable population groups.

Inter-municipal cooperation

Establish cooperation with neighbouring communities and state structures to share experience, solve problems jointly, and attract aid — networking within Ukraine and with European communities.

Investment in infrastructure

Renew and develop infrastructure — roads, residential and public facilities — improving living conditions and creating favourable ground for economic activity and investment growth.

The combined effect

Implemented together, these measures help Group V communities improve their situation, create conditions for business and infrastructure, reduce unemployment, and raise residents' quality of life.

From classification to policy

The results can be used to optimize state regional development policies, design community support programs, and improve funding mechanisms for post-war recovery. The approach can serve as a foundation for further research into the socio-economic development of territorial communities in crisis situations.

References · 23 sources

List of sources used

  1. Handalić Plahonjić M., Kuka F. Sarajevo – Potential and Limitations in the Development of the City and Region. Uprava. 2023. 14(2). pp. 104–118. doi.org/10.53028/1986-6127.2023.14.2.104
  2. Rabe M., Norek T., Widera K., Gawlik A., Łopatka A., Gutowska E. Sustainable Development of the Region – Biomass Potential on Example of the West Pomeranian Region. Journal of Security and Sustainability Issues. 2023. 13(1), pp. 337–344. doi.org/10.47459/jssi.2023.13.35
  3. Neslihan E., Rammah Y., Kamber K., Özpolat Ö. F., Güldüren M., Kırmızıtaş H. Assessment of Wind Energy Potential Across Various Regions of Turkey. dx.doi.org/10.2139/ssrn.4706892
  4. Erfan V. Y., Gazuda S. M., Voloshchuk N. Y. Natural resource potential of the regional spatial and economic system development. Financial and Credit Activity Problems of Theory and Practice. 2018. 1(24). pp. 434–442. doi.org/10.18371/fcaptp.v1i24.128233
  5. Bezama A., Hildebrandt J., Thrän D. Analyzing the Potential Environmental and Socio-Economic Impacts of Regional Energy Integration Scenarios of a Bio-Based Industrial Network. Sustainability. 2022. 14. 15886. doi.org/10.3390/su142315886
  6. Bilous Ya. Yu. Theoretical aspects of organizational and economic support for the development of territorial communities of Ukraine. Efektyvna ekonomika. 2022. No. 3. economy.nayka.com.ua
  7. Dvihun A. Resilience and Potential of Communities in War: Threats and Countermeasures. Review of Transport Economics and Management. 2024. 10(26). pp. 34–38. doi.org/10.15802/rtem2023/300138
  8. Lapishko M., Tiazhkorob I., Lapishko Z., Fedevych L. Trends in the development of financial and investment potential of the territorial economy. Financial and Credit Activity Problems of Theory and Practice. 2021. 2(37). pp. 332–342. doi.org/10.18371/fcaptp.v2i37.230300
  9. Ostrovska H., Andrushkiv B., Tsikh H., Boichyk I., Stavnycha N. Formation of priorities for the development of intellectual potential in the context of a knowledge-based economy. Financial and Credit Activity Problems of Theory and Practice. 2022. 1(42). pp. 415–427. doi.org/10.55643/fcaptp.1.42.2022.3561
  10. Patytska Kh. Financial and economic potential of territorial communities: mechanisms of functioning and activation: monograph. Institute of Regional Research named after M. I. Dolishniy of NAS of Ukraine. Lviv. 2019. 209 p.
  11. Hrynchyshyn I. Socio-economic differentiation of territorial communities in the context of the formation of a new paradigm of regional development. Ekonomika ta suspilstvo. 2020. (22). doi.org/10.32782/2524-0072/2020-22-61
  12. Podzizei O. O., Shmatkovska T. O. Theoretical foundations of restoring the competitiveness of territorial communities and regions. Halytskyi ekonomichnyi visnyk. 2023. No. 82. pp. 12–18. doi.org/10.33108/galicianvisnyk_tntu2023.03.012
  13. Dubovyk A. Innovation and investment potential of territorial communities, ways and tools to increase it. Ekonomika ta suspilstvo. 2024. No. 62. doi.org/10.32782/2524-0072/2024-62-94
  14. Dzvinchuk D. I., Bursa O. I. Human capital as a determining factor in the development of territorial communities in Ukraine. Ekspert: paradyhmy yurydychnykh nauk i derzhavnoho upravlinnia. 2024. No. 1(29). pp. 15–23. doi.org/10.32689/2617-9660-2024-1(29)-15-23
  15. Tomashuk I., Tomashuk I. Strategic foundations of innovative development of territorial communities. Ekonomika ta suspilstvo. 2023. (57). doi.org/10.32782/2524-0072/2023-57-81
  16. Uhodnikova O., Rudachenko O., Troian V., Bogdan N., Kravtsova S. Evaluation and analysis of the socio-economic potential of territorial communities. Multidisciplinary Science Journal. 2024. 7(1), 2025006. doi.org/10.31893/multiscience.2025006
  17. Smachylo V., Dymchenko O., Rudachenko O., Bozhydai I., Khailo Y. Formation of Strategies for the Development of Startup Ecosystems as a Prerequisite for Sustainable Entrepreneurship. In: Semenov A., Yepifanova I., Kajanová J. (eds) Data-Centric Business and Applications. Lecture Notes on Data Engineering and Communications Technologies. 2024. vol 194. Springer, Cham. pp. 1–19. doi.org/10.1007/978-3-031-53984-8_1
  18. Decentralization in Ukraine. decentralization.ua
  19. Pareto V. On the Economic Phenomenon. International Economic Papers. 1953. No. 3. L. P. 184.
  20. Blackman I., Chan E. Using Pareto Principle plus statistic methodology in establishing a cost-estimating model. wbc2013.apps.qut.edu.au
  21. Koch R. The 80/20 Principle. The Secret of Achieving More with Less. KM-Buks. 2022. 400 p.
  22. Sotska Y. I. Methodological Principles of Cluster Analysis of Competitiveness of Ukrainian Banks. Financial and Credit Activity Problems of Theory and Practice. 2015. 2(19). pp. 177–185. doi.org/10.18371/fcaptp.v2i19.57261
  23. Scitovski R., Sabo K., Martínez-Álvarez F., Ungar Š. Cluster Analysis and Applications. Springer Cham. 2021. 271 p. doi.org/10.1007/978-3-030-74552-3