Shifting Sands: Examining and Mapping the Population Structure of Greece Through the Last Three Censuses

Abstract

This paper aims to facilitate a more nuanced understanding of regional disparities in the population age structure at a local scale by applying a recent method for visualizing these disparities. Utilizing data from the three most recent population censuses in Greece, this method applies advanced data visualization techniques to map age distributions, highlighting significant variations in aging patterns across municipalities, towns, and districts. Traditional demographic analysis often overlooks local heterogeneities, leading to broad policies that often fail to address the unique needs of specific regions. Detailed maps are created by integrating geographic data with census data (using R and GIS), enabling policymakers to pinpoint areas with specific demographic challenges and opportunities. This localized approach reveals critical insights, such as regions experiencing rapid population aging, areas with younger population profiles, and zones undergoing demographic transitions. The visualization tool significantly improves the formulation of targeted strategies, enhancing the effectiveness of policies related to healthcare, workforce planning, and social services distribution. Through case studies and comparative analysis, we demonstrate the practical applications and advantages of this method in shaping public policy and strategic planning. This paper contributes to the field of geodemography by introducing and demonstrating a visualization method that enhances the accuracy of demographic analysis, providing policy makers with useful information to better address local demographic challenges and tailor strategies to specific regional needs.

1. Introduction

A country’s population census serves as the primary source of information regarding the population’s size, geographic distribution, and social, demographic, and economic characteristics (UNFPA 2002; Baffour et al. 2013; Killick et al. 2016; MacDonald 2019). The census aims to produce statistics that are relevant and useful to the end users—citizens and organizations (Warren 2014). This is its main statistical objective (Ricciato et al. 2020). Therefore, every action related to the census should focus on delivering a product that meets the needs of these users, whether they are individuals or organizations interested in the data. The data collected through the census offer insights into various aspects of society, enabling the implementation of appropriate policy interventions. By ensuring that census data are effectively utilized and communicated, the development of targeted policies and strategies that better address the diverse needs of society is enhanced.

The organization responsible for conducting the census meticulously plans and designs all of the necessary tools and measures to ensure that the task is successfully executed. Given that the data are ultimately intended for user consumption, many of the questions included in the census may be formulated in consultation with these users. Conducting a census is an expensive undertaking, funded by the users themselves, which necessitates careful and thoughtful design (United Nations 2001).

Censuses are primarily conducted to gather crucial information that governments need to develop appropriate policies, improve public services, and support research by providing accessible data. These data serve as the foundation for understanding demographic trends and making informed decisions at both local and national levels. They can be used to produce statistics on the population and its characteristics, offering a snapshot of the population in a specific area at a specific point in time, complete with spatial references. Traditional censuses, involving the direct enumeration of the entire population, remain widely used and provide invaluable resources for comprehensive and detailed studies of a country’s demographic profile (Yildiz and Smith 2015).

What sets the census apart from other statistical data sources is its comprehensive geographical coverage. This wide-reaching scope allows for a detailed and representative analysis of demographic patterns across diverse areas. It encompasses the entire country, ensuring that every region is included. Another key advantage is the comparability of statistics across different census periods, allowing for consistent and reliable data on all individuals residing in any specific area or subgroup within the population (UNFPA 2002).

Demography is the study of populations (Landry 1945). Population structure is a fundamental indicator of a country’s demographic profile, reflecting trends in birth rates (Tragaki and Lasaridi 2009), mortality rates (Andreopoulos et al. 2021, 2023a, 2023b), and migration (Rovolis and Tragaki 2006; Vollset et al. 2020). Grasping the dynamics of the population structure at both the national and local levels is crucial for policy makers, researchers, and communities, as it offers insights into societal challenges and opportunities (Matthews and Parker 2013). Over the past two decades, Greece has experienced substantial demographic changes, driven by factors such as a rising life expectancy, declining fertility rates, and changing migration patterns. Understanding these demographic shifts at a local level is essential for capturing the nuances of population changes across regions. Mapping and analyzing the population structure in Greece during that period offers critical insights into the country’s demographic evolution and can inform the development of strategies to address emerging challenges. Population aging has emerged as a prominent demographic trend in Greece, mirroring the global patterns observed in many developed countries. As life expectancy rises and birth rates continue to decline, the proportion of older persons within the population is steadily growing (Kotzamanis 2009; Kotzamanis et al. 2018). This demographic shift has significant implications for various aspects of Greek society, including healthcare systems, labor markets, and social welfare programs. Regional disparities in aging patterns require tailored policy responses. To effectively address the consequences of an aging population, it is crucial to understand the spatial dimensions of this phenomenon. In addition to broad national policies, targeted interventions tailored to the specific needs and conditions of each region are essential. Developing and implementing these interventions requires the use of spatial data, employing modern methods of analysis and presentation.

Apart from births and deaths, the population structure of Greece has been shaped by movements, both internal and international (Kotzamanis 2021). Internal migration, mostly driven by economic inequalities and employment opportunities, has led to a strong redistribution of the population within the country, while migratory waves, both incoming and outgoing, have reshaped the country’s demographic composition.

The demographic shifts observed in Greece over the past two decades are not isolated events but are related to broader social, economic, and political developments spanning the last 40 years. Demographic changes occur slowly and their consequences are not immediately reflected by the indicators. However, certain events can accelerate these changes. The economic crisis in the late 2000s, which severely affected the Greek economy and led to austerity measures and structural reforms, had a lasting impact on population dynamics (Kotzamanis et al. 2017). High unemployment rates, especially among young people, have affected migratory flows, with many Greeks seeking opportunities abroad in search of better prospects. In addition, the economic recession had an impact on family planning decisions, contributing to lower fertility rates and changes in household structures.

Besides socio-economic variables, the spatial distribution of the population in Greece is also influenced by environmental factors such as topography, climate, and natural resources. Coastal areas and islands, for example, may attract more inhabitants due to their favorable climate and access to marine resources, leading to a higher population density. This demographic shift is often influenced by regional economic and lifestyle factors. Conversely, mountainous or remote areas may experience depopulation or population aging due to limited economic opportunities and infrastructure. Therefore, environmental assessments are fundamental in understanding the geographic heterogeneity of the population structure in Greece and identifying areas that may be vulnerable to demographic challenges or environmental risks (Kalogeropoulos 2020; Kalogeropoulos et al. 2023).

Mapping and analyzing the population structure in Greece over the last twenty years provides valuable information on the demographic evolution of the country and the challenges and opportunities it presents. Focusing on trends in population aging, migration, and regional disparities, researchers can identify areas that need targeted interventions and inform evidence-based policy decisions. This approach enables a deeper analysis of demographic dynamics at the local level. Understanding the complex interplay of economic, social, cultural, and environmental factors is essential for developing integrated strategies that promote sustainable development and social cohesion across the country. These insights help to inform data-driven decision making at local and national levels. Therefore, mapping census data are another way to disseminate these data (van Elzakker et al. 2003; Kalogeropoulos 2020).

In this paper, a multidisciplinary approach that integrates demographic data using spatial analysis techniques such as R and GIS at the municipality level is proposed (Wardrop et al. 2018). For the analysis and mapping of the population structure in Greece, a new method to capture population aging was used, as proposed by Kashnitsky and Schöley (Kashnitsky and Schöley 2018). This approach maps age structures with a tripartite color coding system, assigning colors to three main age groups: >15, 15–65, >65 years. Each municipality is assigned a specific color according to its age structure. The color coding has been modified to retrieve data from Greek databases, allowing the population structure of each municipality to be visualized accurately. This paper contributes to geodemography by presenting and indicating a visualization technique that boosts the precision of demographic analysis, providing policy makers with valuable information to better address local demographic challenges and adapt policies to specific regional needs. Furthermore, by utilizing data from the three most recent censuses in Greece, this study provides a valuable longitudinal perspective on population aging trends.

2. Methodology

2.1. Data Used

The data used in this paper come from the population censuses conducted by the Hellenic Statistical Authority (ELSTAT) for the years 2001, 2011, and 2021. Specifically, data on the number of people in the age groups 0–14, 15–65, and >65 years at the municipality level were used (for 2001 and 2011, data were reduced to the data for the 332 municipalities present in 2021). The next Figure (Figure 1) presents the study area, Greece (regions of).

2.2. The Used Methodology

For the purposes of studying and mapping the age structure of the Greek population, a new approach was used to investigate the diversity of population aging. This approach was originally proposed in Kashnitsky and Schöley work (Kashnitsky and Schöley 2018). This method has since been adapted for various regional studies. In these studies, the population structure of Europe is investigated at the regional level. Instead of any single summary measure for aging, the entire population age structure is mapped using a tripartite color coding system (palette)—a technique that maximizes the amount of information conveyed by colors (Figure 2). This figure shows the three main categories to be mapped and relates to the age groups <15, 15–65, and >65 years. The colors purple, blue, and yellow are assigned to these groups, respectively. Depending on the population structure, each spatial unit (municipality) is given a specific mixture of the above colors depending on its age structure (each municipality is shown as a point in the palette), as shown in Figure 2. This means that the closer to blue, the greater the relative presence of a working-age population in this municipality. Intermediate shades represent a mix of these age groups. The closer to purple, the younger the population (% of children under 15 years old), and the closer to yellow, the more aged the population of the municipality. All the intermediate colors visualize the remaining percentages of the population structure of each municipality. Each municipality is presented as a black dot within the color palette.

With this approach, each element of a three-dimensional data array composition, such as the three age groups here, is represented by a unique color. This method enables the visualization of complex demographic data in a clear and intuitive way. The use of color mixing to encode multidimensional data into a single feature has been proposed by several authors (Dorling 1991; Kashnitsky et al. 2017; Kashnitsky and Schöley 2018). Since then, the approach has been used to map electoral outcomes in a three-party system, the composition of the labor force by sector, soil textures, the composition of Arctic sea ice cover, and causes of death (Schöley and Willekens 2017). This method has proven to be a versatile way to represent diverse datasets through visual means. In this paper, color coding was used to investigate differences in population structures in Greece by modifying the R code freely provided by Kashnitsky and Schöley (Kashnitsky and Schöley 2018). Thus, while their code was directly linked to the corresponding Eurostat data (descriptive and spatial), in this paper it was modified to retrieve the corresponding data for Greece from indicated databases. This ensures an accurate representation of demographic proportions in the visualizations. The percentages of each category are superimposed on the mapping color palette to give each municipality the corresponding color representing its population structure.

2.3. Software Used

The data analysis was conducted using R software (version 4.1.3). Specific packages used within R included ‘tidyverse’ for data manipulation and visualization, ‘rgdal’ for handling spatial data, and ‘ggplot2’ for creating plots, as provided by Kashnitsky and Schöley (Kashnitsky and Schöley 2018). We avoided using the packages ‘sf’, ‘stars’, and ‘terra’ to prevent potential comparability issues with other Tricolore publications. Then, the maps extracted from R were imported into ArcGIS Pro 3.2 (ESRI, Redlands, CA, USA) to be supplemented with additional data such as a layer representing the regions of Greece.

3. Results and Discussion

Analyzing primary data from the three most recent censuses reveals some key trends in the evolution of the country’s population structure. Notably, the 15–64 age group has experienced an overall decline. This indicates a reduction in the proportion of the population within this working-age group between 2001 and 2021. In contrast, the over 65 age group has shown an overall increase, reflecting a rising number of older adults during the same period. The 0–15 age group presents a more complex picture, with some areas experiencing declines and others seeing increases. However, the general trend suggests a decrease in this younger demographic, as negative changes outnumber positive ones.

Overall, these findings indicate a demographic shift towards a more aged population between 2001 and 2021, characterized by a decline in the working population (15–64 years) and a rise in the older population (over 65 years). In addition, there appears to be a slight decrease in the proportion of the population aged 0–15 years, although the trend in this age group is less consistent.

The following maps eloquently depict the changes in the population structure in Greece overtime, at the municipality level.

The map in Figure 3 illustrates the changes in the country’s population structure following the 2001 census, focusing on the level of municipalities and showing the regional boundaries. The color palette reveals that the average population structure, or intersection, is found in regions with a lower proportion of elderly individuals and a higher proportion of adults.

Additionally, there is a noticeable dynamic in the population structure of Greece’s two major urban centers: the areas surrounding the Municipality of Athens (Athens Metropolitan Center—Attica Region) and the Municipality of Thessaloniki (shaded in purple-blue—Central Macedonia Region). This trend is understandable on a macroeconomic level, as these centers house a significant portion of the country’s active workforce, which drives employment opportunities.

Furthermore, there is a distinct pattern, forming an arc from the municipalities on the northwestern edge of Greece (e.g., Konitsa, Pogoni, Filiates) within the Epirus Region, extending to the westernmost part of the Peloponnese Region (e.g., Pylos-Nestoros and Messini). These areas are characterized by a notably aging population (over 65 years old), with exceptions seen in the capitals and larger urban centers of the respective regional units, such as Ioannina (R.U. Ioannina–Epirus Region), Agrinio (R.U. Agrinio–Central Greece Region), Patras (R.U. Achaia–Western Greece Region), and Kalamata (R.U. Messinia–Peloponnese Region).

The map in Figure 4 shows the population structure of the country after the 2011 population census.

Figure 4 displays population data similar to that in Figure 3 but focuses on the aged population, highlighted in yellow shades, which is primarily located in Western Greece. This map clearly illustrates the shift towards a higher proportion of individuals aged 65 and over. Out of 332 municipalities, only 25 have experienced a decrease in this age group, while the remaining municipalities have seen increases. Notable examples include the municipalities of Amphipolis (Central Macedonia Region), Nestoriοu (Western Macedonia Region, and Visaltia (Central Macedonia Region), which have experienced increases of over 10%, and Andritsaina-Krestenon (Western Greece Region), Sintiki (Central Macedonia Region), Central Tzoumerka (Epirus Region), and Agrafa (Central Greece Region), where the increase in the aged population exceeds 9%.

The following figure, Figure 5, depicts the population structure of the country based on the 2021 census.

As shown on the map above, there is a clear shift towards the above 65 years of age group. The data reveal a consistent pattern: the population aging observed between 2001 and 2011 is mirrored in the data from the 2011 and 2021 censuses. The latest census continues to reflect the trend in population aging observed over the last 20 years, mainly in Western Greece. The data reveal a consistent pattern: decline in the younger age group (0–15 years), minimal changes in the working-age population, and a noticeable increase in the elderly (over 65 years). This trend is evident throughout Greece and highlights broader demographic shifts, such as falling birth rates and an aging population. Specifically, cities like Komotini and Alexandroupoli (Eastern Macedonia and Thrace Region) in eastern Greece have seen relatively small reductions of 1% to 3% in the younger population. These decreases are often accompanied by a modest rise in the elderly population. For instance, the older people living in Komotini and Alexandroupoli rose by almost 3–4%. On the other hand, there are also many municipalities where changes are far more pronounced, e.g., Grevena (Western Macedonia Region), Dodoni and Zagorio (Epirus Region). Huge drops in the 15–65 age group (10% or more) and large increases in those over sixty are prominent in these areas. Depopulation, due to the outward migration of the younger generations, has aggravated the rapid aging process. For example, in Arriana, the older population grew by 9.82%, while in Iasmos the above increase reached 11.60% (Eastern Macedonia and Thrace Region). Thus, this shift in demographics will require major changes to public policy, healthcare, housing and accommodation, and social services for the elderly.

Examining the whole period over the last 20 years (three censuses), the 0–15 age group decreased in almost all of the municipalities, which raises concerns about the future population and economic health. Municipalities like Myki (Eastern Macedonia and Thrace Region) and Aminteo (Western Macedonia Region) saw big drops of 8.58% and 10.28%. This could directly affect the quality and quantity of services provided to young persons (local schools, youth programs, and jobs) causing a vicious cycle that further aggravates depopulation and aging.

The above figures show changing population patterns that, if not taken into consideration, may undermine regional development. The decrease in young and working-age people, along with more old people, could make it hard to keep the economy going and provide social services. Local leaders might need to come up with new ideas to deal with these changes. They might try to bring in more young people or increase the number of babies being born, while also helping older people more.

This is, of course, a consequence of the decrease in birth rates, the flight of a large part of the economically active population in the years of the economic recession and vice versa, the increase in the survival rate of elderly people, etc.

Due to the geographical nature of census data, maps are a critical element in conducting a census or survey, but also in the dissemination of the results. Therefore, mapping is a very important function of statistical services. Maps are used both by census takers, at the survey-recording stage of data collection, and in print and electronic publications, as a supplement to statistical data (Bower 2010).

The decision to adopt this particular color-coding method, originally proposed by Kashnitsky and Schöley, was due to its ability to visually capture the complexity of the age structure of the population in a single, intuitive image. This approach allows for a holistic longitudinal comparison of municipalities. In comparison with other methods, such as spatial maps, the tripartite palette approach provides information not only on age groups but also on the balance between them. Given the aging population in Greece, the method offers clarity and depth in its perspective, thus facilitating policy making. Alternative ways of presenting the population structure would require many more maps in order to achieve the same level of detail. However, the chosen technique compresses the multidimensional age data into a single visual effect.

Kashnitsky and Schöley’s R code was modified in order to achieve compatibility with Greek demographic data. The original code analyzed data from Eurostat databases. This link was modified so that the data were extracted from the database created for the needs of the study. Through testing, the way in which the spatial data were extracted was also modified to ensure that the tripartite mapping of the population structure by municipality was accurately mapped.

Traditionally, the role of maps in a census has been to support the enumeration and presentation of residents and dwellings at all stages of the census process (before, during and after the census) and to present the results of this on a map. The rapid technological development that has taken place in recent years (both in GIS science and cartography) has significantly expanded this role. In addition to the more efficient production of population census maps and thematic census maps, GIS now plays a key role in familiarizing society with the census process and in the more thorough analysis of the data it provides (United Nations 2000). The cartographic presentation of census results is a powerful tool as it can support the identification of local patterns of important demographic and social indicators. As such, maps are an integral part of management policies in the public and private sectors. Moreover, highlighting spatial variations is important, as it is imperative and necessary in economic and social planning on the part of the state at local and regional levels (Kotzamanis and Pappas 2000; Pappas 2001).

4. Conclusions

The present study investigated the evolution of population structures in Greece during the last two decades, based on data from population censuses. The main emphasis was placed on mapping and analyzing population aging at the local level, specifically within municipalities. To achieve this objective, the study used a new approach that adopts a three-color coding/grading system to visually represent the age structure in each municipality.

The findings reveal notable variations in population aging across different regions of the country. Some municipalities have already experienced higher levels of aging, while in others aging is progressing more rapidly compared to the national average. These disparities may be influenced by factors such as regional economic conditions, migration trends or access to healthcare services. The use of tripartite color coding has proved particularly effective in visualizing these differences, making it easier to identify areas that require targeted attention.

The study employed a relatively new methodological approach that utilizes trivariate color coding to represent age structures. This technique has been highly effective in depicting complex population structure data and highlighting spatial variations. By using color coding to illustrate three distinct age groups (0–14, 15–65, and over 65 years), the method presents demographic information in a clear and visually engaging manner, enhancing our understanding of local demographic trends.

The insights provided by the study are valuable for policymakers, researchers, and local communities, aiding in the development of targeted strategies to address the challenges of an aging population. Additionally, the tripartite color-coding methodology can be applied to other demographic or socio-economic data, facilitating the visualization of complex information and the identification of spatial patterns. The study provides a lot of information on demographic developments and variations in Greece. Firstly, the visualization technique used revealed distinct patterns of population aging at a very detailed level, thus highlighting the municipalities experiencing the fastest demographic changes. Conversely, municipalities with younger populations are highlighted, suggesting emerging opportunities for investment in youth-oriented education and infrastructure. In addition, the analysis reveals those transition zones where demographic profiles are changing, indicating areas that may need to be targeted for the implementation of targeted policies. Mapping using this method therefore provides the necessary knowledge for the implementation of appropriate policy interventions.

Despite the valuable insights provided by this study, there are important limitations that need to be considered. One of these is the reliance on data from only the three most recent censuses. This means that long-term trends in the population structure may not be detectable. The method used implies the use of detailed data. This means that it should be used according to the number of spatial entities used at any one time. Finally, since the input data are only population data, it is assumed that the socio-economic conditions of each region are not taken into account and therefore these should be used in the discussion of the results.