En los últimos años, las personas jóvenes han enfrentado un empeoramiento de su situación económica en comparación con otros grupos de edad, impulsado por factores como el aumento de la precariedad laboral, el encarecimiento de la vivienda y mayores tasas de desempleo. Como consecuencia, presentan ahora una mayor incidencia e intensidad de la pobreza que en periodos anteriores. Este estudio analiza el papel de la edad en la distribución de la renta y la pobreza monetaria en Galicia, centrándose en la evolución de la renta per cápita, la desigualdad de renta, la contribución de las diferentes fuentes de renta y de la pobreza durante el periodo 2007-2021. El análisis distingue entre tres grupos de edad: jóvenes (<35 años), personas de mediana edad (35-64 años) y personas mayores (> 65 años). Los resultados indican que las personas jóvenes se han desplazado progresivamente hacia los tramos inferiores de la distribución de la renta, con un aumento de las tasas de pobreza, en contraste con las mejoras observadas en los otros dos grupos de edad a lo largo del tiempo. Además, los factores que influyen en la distribución de la renta y la pobreza varían entre grupos de edad y han evolucionado durante el periodo estudiado. Estos resultados subrayan la necesidad urgente de políticas específicas para abordar estas desigualdades y fomentar la cohesión social.
The substantial rise in income inequality emerges as one of the most pressing global challenges, paralleling the climate crisis (). This issue's significance is reflected with its inclusion in the Sustainable Development Goals of the 2030 Agenda (SDGs 5 and 10), which aims at diminishing inequalities and advance in gender equality. However, the Great Recession and the Covid-19 pandemic have acted as a setback to the achievement of these goals in some societies.
In the case of Galicia (Northwest Spain), during this period, there has been a significant increase in inequalities and poverty (; ). Among the reasons for this increase, the worsening of well-being for the younger population plays a central role, which, however, has been addressed in a limited way by the existing literature. Given the above, it is essential to quantify and study the changes in income distribution and poverty according to individuals' age in order to improve the design of public policies aimed at supporting the most vulnerable.
Galicia represents a particularly interesting case study due to its status as a peripheral region within Spain and its demographic profile, characterised by a deep decline and demographic ageing, which has led to profound demographic, social and economic adjustments (). Given the strong demographic ageing, Galicia presents the ideal context to study how inequalities between age groups have evolved in recent years. As well as to make a first approximation to the possible effects of the advance towards a society in which gerontocracy prevails.
The main objective of this paper is to study income distribution and poverty in Galicia across age groups during the 2007–2021 period. More specifically, this research aims to analyse how income inequality and poverty incidence have varied among different age groups and to identify the determinants of income distribution and poverty and the role of the different income sources and social benefits. In this regard, Galicia serves as a highly relevant case study due to its status as a peripheral region and its pronounced demographic ageing. To conduct this study, data from the Income and Living Conditions Survey is used, and the population is divided into three age groups: young people (<35 years), middle-aged individuals (35–64 years), and older adults (>64 years). A descriptive analysis is carried out alongside machine learning techniques, specifically random forests.
This paper is divided into five sections, including the present introduction. The second section provides a literature review on the role of age in inequality and poverty in Galicia and within the Spanish context, as well as the evolution of inequality and poverty in Galicia. The third section outlines the data sources and methodologies used for the empirical analysis. The fourth section presents and discusses the results obtained. Finally, the main conclusions of the research and the resulting recommendations for economic policy are presented.
This section provides a review of the literature on the role of age in income distribution and poverty, as well as on the evolution of income distribution and poverty in Galicia. First, the role of the generational gap in income distribution and poverty is analysed in the Spanish and European context. Second, a review of studies and reports on the evolution of income distribution and poverty in Galicia is conducted, aiming to identify the most relevant trends and the role played by the generational gap in these dynamics.
Economic inequality is one of the topics under discussion in contemporary literature, highlighted for its impact on economic growth, social cohesion, and political stability (; ; ). Inequalities manifest in various forms, including income, wealth, access to opportunities, and educational outcomes. These disparities have increased due to factors such as technological change and recurring economic crises, which have intensified the focus on the most vulnerable groups in society ().
A relevant perspective for analysing this phenomenon is intergenerational inequality. Generational gaps can be defined as structural and systemic disparities in access to resources, wealth, and opportunities between different age cohorts, resulting from specific economic, social, political, and demographic dynamics. These gaps pose a central challenge for contemporary economies, particularly in societies with aging demographic structures and marked by recurrent crises ().
In the European context, the increasing precariousness of labour, limited access to housing, and income distribution inequality have placed younger generations in an economically vulnerable position (; ). These disparities are particularly concerning in peripheral regions like Galicia, where factors such as population aging and reliance on traditional economic sectors may exacerbate inequalities between age groups (; ; ).
The generational gap is a multifactorial phenomenon, meaning its existence and evolution depend on various social, economic, and political factors. Moreover, the explanatory factors behind generational gaps can be attributed to both cyclical and structural causes. Existing literature on generational gaps in the European and Spanish contexts identifies three key groups of factors as the most relevant: labour precariousness and economic crises, public policies and demographic ageing, and the housing market and access to housing. The role of each of these factors is addressed separately below, although clear connections exist between them.
Labour precariousness is particularly impactful on youth, affecting them from their entry into the labour market, coupled with higher unemployment rates (). define precarious employment as a situation where work fails to provide employees with the security of a minimum decent standard of living. Similarly, and consider precariousness as a lack of protection, insecurity, economic or social vulnerability, and instability. describe precariousness as the absence of stable employment, flexibility, temporary contracts, low wages, and lack of control at work, which prevent the consolidation of sufficient income and stability to establish a future and a dignified life.
An immediate economic consequence of labour precariousness is the risk of poverty, understood as the lack of resources or income (; ). Precariousness is closely linked to poverty, as some precarious workers are also poor. According to , these situations of exclusion are rooted in a combination of vulnerabilities that merge low-income levels with other disadvantages, such as housing problems or lack of access to basic educational or healthcare resources. This concatenation of circumstances leads to a precariousness trap, whereby individuals in precarious conditions and relying on subsidies may not be encouraged to participate in a labour market that offers no guarantees. Furthermore, should they rapidly exit these conditions, they may have to incur transaction costs again to regain access to minimum income subsidies. In the Spanish context, young people (<35 years) have progressively shifted towards the lower levels of the income distribution, while older adults (≥65 years) have experienced improvements thanks to a robust social protection system (; ). However, these phenomena have been addressed only to a limited extent in the literature, particularly regarding income sources and their role in wealth distribution.
The vulnerability of younger generations may be further compromised by the more significant impact of economic crises (). For instance, highlights that the Great Recession (2007–2009) disproportionately affected young people and working-age cohorts. indicate that, as a consequence of the crisis, younger generations not only experienced reduced incomes but also lost future opportunities. This idea is grounded in the notion that current harm may create obstacles to future recovery, with long-term negative effects ().
Another factor that differentially affects generational cohorts is access to public resources. This may be due to the social policy infrastructure, where historically public transfers have favoured older generations, perpetuating intergenerational inequality (). and identify that in Spain, the fiscal system has been ineffective in mitigating intergenerational inequalities, favouring older generations in terms of social benefits. This limits the redistributive capacity of the state.
Intergenerational inequality is a complex phenomenon that captures disparities in access to resources, wealth, and opportunities among different age groups. highlight that this form of inequality is intrinsically linked to demographic structures and welfare policies. Specifically, they emphasise that pension systems and wealth accumulation through property ownership have favoured older adults to the detriment of younger generations. Additionally, introduce an Intergenerational Inequality Index (IGI) to quantify these disparities, demonstrating that younger generations face greater disadvantages in terms of income and access to resources. Their analysis underscores how public transfer policies, rather than mitigating these inequalities, often perpetuate them. Furthermore, as highlighted by , the demographic ageing experienced in certain countries and regions has increased pressure on welfare systems, further prioritising older generations over younger ones.
Regarding recent public policies in Spain, measures such as the Minimum Living Income (Ingreso Mínimo Vital, IMV) have shown limited impact on young people, who are often less likely to benefit from programmes aimed at combating extreme poverty. In contrast, older adults have been the primary recipients of these benefits, further strengthening their position at the upper end of the income distribution (). However, the lack of a comprehensive approach to addressing social exclusion among young people, particularly in impoverished rural and urban areas, remains a critical challenge. highlights the need for holistic interventions that combine access to affordable housing, youth employment programmes, and participation in community networks as essential components for reducing exclusion.
identifies an increase in inequalities both between young people and adults and among youth themselves. In addition to the differential impact of crises, homeownership is another determinant of inequality between adults and young people. emphasize that young households face greater difficulties in acquiring housing, thereby increasing their socioeconomic precariousness. The reasons behind the greater difficulty in accessing housing among young people stem from both the deterioration of their labour market income and the rising cost of housing, as well as stricter credit conditions (; ).
provides an in-depth analysis of changes in housing access and the real estate market in Spain since the Great Recession, concluding that Spain is moving away from the traditional Southern European residential model characterised by high residential stability. This shift has resulted in an increase in rental housing across most social strata, particularly in major metropolitan areas. However, as highlights, this rise in residential instability has disproportionately affected young adults. Fewer than 70% of individuals aged 30 to 34 have achieved emancipation, a trend accompanied by a significant increase in rental prices. As a result, 50% of those under 35 who rent their homes face a severe financial burden to afford their housing costs.
Studies on income distribution and poverty in Galicia have been limited in the existing literature, with a few standing out such as , , or . However, these investigations provide a general overview of the evolution of inequality and poverty without delving into the role of the generational gap.
and offers a historical overview of income distribution in Galicia from 1973 to 1991, situating these trends within the broader context of Spanish economic developments. The results demonstrate a general decline in both inequality and polarization over the course of the period under examination, with particularly noteworthy improvements observed in the 1980s. The changes are attributed to economic growth and shifts in population distribution across income groups, thereby suggesting a movement towards greater income equality and reduced polarisation.
examine the impact of the economic crisis on income distribution in Galicia between 2007 and 2010, noting an increase in inequality and poverty, particularly affecting the younger population in the Atlantic areas due to higher unemployment rates. The study employs a range of measures, including the Gini, Atkinson, and Theil indices, to illustrate that the crisis has intensified income disparities, resulting in a shift in the poverty map from older, pension-reliant interior regions to younger, unemployment-affected coastal areas.
analyse the initial effects of Covid-19 pandemic in the income distribution in Galicia. The study shows that the pandemic increases the income inequalities in primary income distribution and that young people have been the most adversely affected by the pandemic, starting from an already disadvantaged position. Additionally, show that social benefits have mitigated income loss for middle-aged and older groups but not for younger individuals.
The AROPE (At Risk of Poverty or Social Exclusion) reports from 2018 to 2022 demonstrate a persistent issue of poverty and social exclusion, with slight fluctuations over the years. In 2018, the AROPE rate was 24%, equating to approximately 652,000 individuals at risk. This rate increased to 25.2% in 2021, with 678,000 people at risk (). Furthermore, the highest AROPE rates are consistently observed among children and adolescents (under 18 years of age), followed by adults (18-64 years of age) and seniors (65 years of age and above). In 2021, the AROPE rate for children was 34%, which was significantly higher than the rates for adults (25.4%) and seniors (19%). This suggests that younger populations are more susceptible to poverty and social exclusion. However, the age group division used does not allow for distinguishing between the situations of young adults and middle-aged adults, as it combines a broad age range (18–64 years) with significant internal heterogeneity.
In the European context, Spain is one of the countries with the highest levels of poverty and social exclusion. Specifically, in 2023 it was in fourth place, behind only Bulgaria, Romania and Greece (see Figure 1). This reinforces the interest of studying the case of Spain.
Other studies examine phenomena related to income inequality and poverty in Galicia that are relevant to understanding the changes brought about by the Great Recession and the Covid-19 pandemic. analyse the influence of tourism on income distribution in Galicia. The authors conclude that, while tourism contributes to economic growth, it also increases income inequality. High-income households benefit more from tourism-related activities, thereby exacerbating existing disparities.
examines the effects of the Great Recession and the initial impacts of the Covid-19 pandemic on employment in Galicia, focusing on age and gender disparities, as well as the structural and sectoral factors that drove job losses. The results show that young workers (16-29 years old) acted as a buffering workforce in both crises: they suffered disproportionate job losses during recessionary phases and experienced sharp recoveries during recovery periods. As the authors point out, this is due to the high prevalence of young workers in temporary jobs, which makes them more vulnerable.
The authors also show how the Expediente de Regulación Temporal de Empleo (ERTE) scheme implemented by the Spanish government helped mitigate job losses during the pandemic by providing subsidies for temporary suspensions or reductions in working hours. However, young workers benefited less from this policy due to the non-renewal of temporary contracts before its implementation.
The studies analysed in this subsection show how Galicia has experienced significant changes in income distribution and poverty in recent decades. Additionally, the generational inequality is highlighted in some of these studies (; ; ; ) as a key aspect in understanding the changes experienced by Galician society in recent decades. However, these studies do not directly address the quantification and determinants of the generational gap in Galicia with sufficient depth. The following sections of this article conduct such a study, aiming to fill this gap in the literature.
This section discusses the main data sources and methodologies used to analyse income distribution and poverty in Galicia, focusing on inequalities between different age groups. First, the data source utilised, and the key variables used in the study are presented. Next, the main methodologies employed to conduct the analysis are outlined.
To analyse changes in income distribution and poverty in Galicia, microdata from the general module of the Income and Living Conditions Survey (Encuesta de Condiciones de Vida) for individuals from households residing in Galicia are used, covering the years 2008, 2011, 2015, 2020, 2021, and 2022 (). The selected editions provide data on the households and individuals surveyed for the previous year, namely 2007, 2010, 2014, 2019, 2020, and 2021. The selection of these years is driven by the need to observe changes generated by the economic cycles that occurred in Galicia (and, in general, in Spain) during this period. More specifically, the inclusion of the years 2007 and 2014 allows for the analysis of changes resulting from the Great Recession by comparing the situation prior to it (2007) and immediately before the start of the economic recovery (2014). The inclusion of the year 2019 enables the analysis of the situation after the economic recovery, before it was interrupted by the Covid-19 pandemic (2020–2021).
The Income and Living Conditions Survey provides information on the income received by households and individuals residing in Spain, as well as details about their specific characteristics and circumstances. Among the characteristics collected are the region in which the household's main dwelling is located and the age of its members, making it an excellent choice for conducting this research. In addition to these advantages, the Income and Living Conditions Survey has a key advantage over other alternatives, such as the Structural Household Survey (Enquisa Estrutural a Fogares) (), in that it offers a greater level of income disaggregation based on its source. Furthermore, the survey microdata collects a sample of the Spanish population and provides a weighting factor that enables representative results for the national and regional population, both for households and individuals. In this research, this weighting factor is applied at the individual level.
In this research, the individuals in the population sample of adults (16 years or older, as defined by the Income and Living Conditions Survey) are divided into three groups based on their age: young (<35 years), middle-aged (35–64 years), and older adults (>64 years). The selection of these groups is based on the need to observe individuals' circumstances according to their life stage. Young adults are defined as those in their formative years and/or in the early stages of entering the labour market, following the criteria of other studies (). The middle-aged group consists of adults who are not classified as young but are of working age. Finally, the older adults group comprises individuals who were at/or above the legal retirement age at the beginning of the period. It is worth noting that this retirement age has varied over the period due to legal reforms.
Within the Living Conditions Survey, some income variables are collected at the household level, while others are recorded at the individual level. Since this research is conducted at the individual level, household-level incomes are evenly divided among the adult members of the household. Thus, an individual’s income consists of both their directly received income and a proportional share of their household's income. Additionally, only net incomes are used.
The fact that some sources of income are not available at the individual level and have to be divided equally may be a limitation when carrying out the analysis by age. In this respect, it is worth highlighting the case of real estate income when young people return to the family home or stay there for a longer period of time. In this case, young people benefit to some extent from imputed rent, even if it is provided by their parents, the worse their individual situation is, the more so. This could lead to the result that real estates rents have become more important for young people. This points to the need for further detailed studies on the role of real estate rents. Table 1 presents the income variables that make up each individual's disposable income and the level at which they are recorded within the Income and Living Conditions Survey.
Source: The authors from .
Regarding the study of the determinants of poverty and income distribution, beyond age itself, this research examines individuals' characteristics related to their gender, type of municipality, highest level of education attained, housing tenure status, employment situation, and occupation. Table 2 presents the variables used to analyse these characteristics.
Source: The authors from .
Using the variables presented in Tables 1 and 2, a study of income distribution and poverty is conducted both overall and for each age group. The income sources listed in Table 1 allow for the decomposition of individuals' income, while the characteristics in Table 2 enable an analysis of the factors that explain individuals' positions in society according to their age. The following section presents the indicators and methodologies used.
This section introduces the indicators used to study changes in income distribution and poverty by individuals' age, as well as the machine learning algorithm employed to analyse the changes in the determinants of income distribution and poverty within each age group. Starting with the indicators, this research uses four key measures: average disposable personal income, the Gini coefficient, the contribution of each income source to disposable personal income, and the poverty incidence rate.
To calculate disposable personal income, which is used in the subsequent indicators, the income sources listed in Table 1 are included. Notably, imputed rent is incorporated into individuals' income, allowing for an analysis of the role of housing as a guarantor of well-being. Including imputed rent as a separate income source from property income enables a distinction between housing's role as a provider of in-kind income and monetary income. Social benefits are also included with some disaggregation to highlight their differing roles depending on their nature. The first type, referred to as unemployment benefits, encompasses those received by unemployed individuals eligible for such support. Retirement benefits include those arising from inactivity due to retirement, illness, or other forms of incapacity. Finally, other benefits refer to aids or allowances aimed at addressing vulnerability or poverty.
The average disposable personal income is calculated for the entire sample of adult individuals and for the different age groups analysed, using their individual weighting factors.
To analyse the evolution of inequality for society as a whole and across the different age groups, the Gini index is used using their individual weighting factors. This measure provides an overall view of changes in income distribution, focusing primarily on shifts in the central part of the distribution.
To analyse changes in poverty across age groups, the percentage of the population at risk of monetary poverty is calculated for the overall population and for each age group. The poverty threshold is set at 60% of the median income and the dummy variable “poverty” indicates whether the individual is in poverty or at risk of poverty (1) or not (0).
To analyse changes in individuals’ positions by age, the distribution of each age group across the deciles of the overall population is calculated. Specifically, income for each population decile is determined for the entire population (the categorical variable “decile” indicates to which population decile the individual belongs; 1-10), and then the percentage of individuals from each age group within each decile is observed relative to the total number of individuals in that group.
To analyse the differing roles played by various income sources, the share of each source in total disposable personal income is calculated. This is achieved by decomposing income into the categories listed in Table 1. Other income sources are excluded due to its marginal role in the individual’s income.
To analyse the determinants of income distribution and poverty within the studied age groups, the variables described in Table 2 are used. These include information on key factors that influence the income level of each individual and, by extension, monetary poverty. To examine the role of each variable, a random decision forest , ) is employed, which helps identify the most relevant variables for explaining the classification of individuals from each age group across different deciles and whether they are at risk of poverty.
Random forest is a supervised non-parametric machine learning methodology used for regression and classification tasks. This approach offers several advantages over other methods, such as logistic or probabilistic models. The main benefit is its non-parametric nature, which does not assume a linear relationship between the determinants of income distribution and poverty (; ).
The functioning of Random Forest allows for the analysis of the importance of the variables used as regressors to explain the classification of individuals into different groups (). Specifically, as analysed in this research, this methodology measures the loss in classification accuracy when each variable is excluded from the regression while keeping the others. This approach enables the identification of the most determining factors for classifying individuals across income deciles and/or into the categories of poor or at risk of poverty versus non-poor.
In this research, the Random Forest methodology is employed to analyse the classification of individuals by age group. The aim is to study the extent to which the determinants of income distribution and poverty vary between individuals of different ages and whether there have been substantial changes within each group over the period analysed. To achieve this, equation (1) is used in the case of income deciles and equation (2) is used in the case of poverty, and 5,000 decision trees are calculated for each equation and age group, where i denotes the individual, a denotes the age group and t denotes the year (2007, 2010, 2014, 2019, 2020 and 2021):
The Random Forest methodology is a supervised machine learning technique used for classification or regression problems. In the study of income or poverty, it enables the identification of the factors that determine which income decile an individual belongs to, or whether they are in a situation of poverty. It is based on the construction of multiple decision trees from random samples of the dataset. One of its main strengths lies in its ability to estimate the importance of each explanatory variable. In this way, it helps to identify which characteristics have the greatest influence on classifying an individual into a specific income decile or poverty status. Furthermore, Random Forest can capture complex interactions and non-linear relationships between variables, giving it an advantage over simpler methods in contexts with heterogeneous socio-economic data.
The way the analysis is carried out also allows to assess the extent to which removing a variable from the classification tree reduces the model’s ability to correctly classify individuals. In the case study, since estimates are made for different years, this methodology also makes it possible to compare which factors have gained or lost influence over the period analysed in the different age groups.
To evaluate the performance of the models, the percentage of correctly classified individuals out of the total number of individuals in each group and year is used. As shown in Table 3, the results are much more accurate in explaining the incidence of poverty compared to explaining income distribution. This is largely due to the fact that explaining poverty incidence is simpler, as it involves a dichotomous variable, whereas explaining a categorical variable with 10 levels (one for each income decile) is inherently more complex.
It can also be observed that the analysed determinants show higher accuracy, both for income distribution and poverty, in the young group compared to the middle-aged and older groups. Additionally, there are significant differences between the middle-aged and older groups, with the former displaying noticeably greater accuracy than the latter.
Another noteworthy aspect is that the determinants appear to lose accuracy in explaining income decile among middle-aged and older households but not among young individuals. In the case of poverty, the determinants maintain a consistent explanatory capacity throughout the period, ranging between 85–90%.
Source: The authors from .
To assess the relevance of each explanatory variable, the variable importance metric based on the Mean Decrease in Gini is used. Each time a variable is used to split a node in a tree, the reduction in impurity achieved at that node is calculated. The importance of a variable is estimated as the sum of all impurity reductions it generates across all nodes where it appears, averaged over all trees in the Random Forest (; ).
This section presents and discusses the results obtained for Galicia following the methodology outlined in the previous section. First, the results related to the evolution of income, inequality measured by the Gini coefficient, and the incidence of poverty are presented for Galicia as a whole and for the different age groups analysed. Second, the evolution of the distribution of individuals across income deciles is displayed using a heatmap. Third, the weight of different income sources within the various age groups and their evolution over the period is examined. Finally, the role of the determinants of income distribution and poverty is analysed through the results obtained using the random forest methodology.
During the 2007–2021 period, economic cycles had a significant impact on individuals’ average income (Figure 2). This is reflected in the evolution of average income (in constant terms), which shows declines during the Great Recession (2007–2014) and the onset of the pandemic (2020) and increases during phases of economic recovery (2014–2019 and 2021). However, distinct patterns emerge when examining the trends for each age group. For young individuals, there is a clear downward trend over the entire period, driven primarily by a sharp decline during the Great Recession (around €4,000 less income in 2014 compared to 2007), a modest increase during the recovery phases, and a less severe decline during the pandemic. For middle-aged adults, the overall trend is upward for the period, although their average income decreased during the Great Recession and the pandemic. For older adults, a consistent upward trend is observed throughout the period.
The evolution observed over these years reveals a clear consequence: young people have experienced a worsening of their situation compared to the other two groups, according to . Furthermore, older adults have seen the greatest improvement in their circumstances, moving from below the average income at the start of the period to above it by the end ().
The decline experienced by young people can be attributed mainly to two factors: reduced access to income sources due to limited access to the labour market, home ownership, and other opportunities typically associated with entering adulthood, or a decline in their earnings as a result of the economic crises. To delve deeper into this aspect, it is necessary to examine the internal inequality among young people and their relative position within the overall income distribution in society.
Figure 3 shows the evolution of the Gini coefficient during the 2007–2021 period for the Galician economy as a whole and for the analysed age groups. The results again reveal a divergent evolution among age groups. On the one hand, young individuals show a clear increase in inequality over the entire period. On the other hand, middle-aged and older individuals display a reduction in inequality throughout the same period.
A notable aspect is the contrasting patterns of inequality observed among age groups during the analysed period. For young and middle-aged adults, inequality increases significantly during the Great Recession and the pandemic, driven by heightened unemployment and precarious working conditions, as highlighted by and . Conversely, these groups experience a reduction in inequality during economic recovery phases as labour market conditions improve. In contrast, older adults exhibit a different pattern: inequality rises during recovery periods, likely due to heterogeneity in pension benefits and investment income, while it decreases during crises, reflecting the stabilizing role of robust social protection systems, as suggested by and .
Young people have not only experienced a decline in their average income and an increase in income inequality among themselves, but as shown in Figure 4, they have also seen a considerable rise in their poverty levels. Additionally, monetary poverty (individuals earning less than 60% of the median income) has increased significantly among young people reflecting their higher exposure to precarious employment and limited access to stable income sources (). This contrasts sharply with the situation of middle-aged and older individuals, who have much lower poverty rates and exhibit a declining poverty trend over the period.
To better understand the changes in individuals' positions, it is necessary to go beyond the Gini coefficient. Figure 5 presents a heatmap showing the distribution of each age group across income deciles for the total population. For instance, the fact that in 2007, 18.78% of young individuals were in the first income decile illustrates that nearly 2 out of 10 were among the poorest 10% of the population.
The results reveal a clear trend towards the concentration of young individuals in the lower income deciles, to the detriment of the middle and, especially, upper deciles. For instance, in 2007, just over 30% of young individuals were in the bottom two deciles, while by 2021, this figure had risen to nearly 50%. Conversely, in 2007, just over 41% of young individuals were above the median income, compared to slightly over 25% in 2021. Another dynamic revealed by the results is the trend towards a greater presence of middle-aged individuals in the middle-income deciles (3–6), although they continue to have a stronger representation in the upper deciles. Finally, older individuals exhibit a trend towards a greater presence in the upper deciles (7–10), although they remain predominantly concentrated in the middle deciles (3–6).
Source: The authors from .
In summary, the trends shown in Figure 3 depict a society where young people dominate the lower classes, middle-aged individuals occupy the middle-upper classes, and older individuals are concentrated in the middle-lower classes. However, there are tendencies towards a further decline for young people and a rebalancing between middle-aged adults and older individuals over the period.
To understand the changes in the situation of different age groups, it is essential to examine the evolution of their income composition. Figure 6 illustrates the trends in the income components analysed (see Table 1) for young individuals, using an index to account for their varying weights. Additionally, Table A.1 in the Appendix presents the absolute values for this group. The results indicate a significant increase in other social benefits (which multiply by more than six), primarily comprising social exclusion assistance. A notable decline is observed in financial income, mixed income, and wages. Conversely, there is a more moderate increase in imputed rent, unemployment benefits, and property income. However, in some cases, these variations may be influenced by how the income is assigned to individuals.
Figure 7 illustrates the changes in income composition for middle-aged individuals (detailed data are provided in Table A.1 in the Appendix). A significant increase in other social benefits can be observed, although this rise is smaller than that experienced by young people. In contrast, there is a notable decline in financial and mixed incomes, as well as smaller reductions in retirement benefits and imputed rent. Additionally, a moderate increase in the share of unemployment benefits, wages, and property income can be observed.
Figure 8 shows the evolution of income composition for older adults (detailed data are provided in Table A.1 in the Appendix). Once again, there is a significant increase in other social benefits, which in this case exceeds that of the other two groups. Notably, there is a sharp decline in mixed income, which reaches clearly negative values in 2021. This anomalous behaviour, not observed in the other age groups, may reflect a specific profile of activities among older individuals particularly affected during the pandemic. There is also a notable increase in unemployment benefits and wages, which is logical given the rise in the legal retirement age introduced during the Great Recession. Conversely, financial income has declined considerably, as seen in the other two groups. Finally, a group of income sources, including imputed rent, retirement benefits, and real estate income, remains relatively stable throughout the period.
The results of the income composition analysis reveal that the category of other social benefits has increased its share across all age groups and, by extension, within society as a whole. This growth reflects the implementation of new aid schemes aimed at mitigating the effects of the Great Recession and the pandemic on the most vulnerable groups. Of particular interest is the sharp increase observed in 2021, likely reflecting the impact of the Minimum Living Income (Ingreso Mínimo Vital) introduced the previous year. However, as the results indicate, these policies are far from being youth-oriented and have disproportionately benefited older individuals. This is noteworthy given that older adults are the age group with the greatest income gains during the period.
Figure 9 illustrates the variation in the weight of different income sources between 2007 and 2021 for the three age groups analysed. The results clearly show that the group with the largest variations is the one formed by young individuals, with these changes primarily due to the decline in the share of wages and the increase in imputed rent. On the one hand, wages represented 51.23% of young people's income in 2007, compared to 35.28% in 2021. On the other hand, imputed rent accounted for 34.11% of young people's income in 2007, in contrast to 49.01% in 2021. These results highlight how the Great Recession and the pandemic have particularly affected young individuals due to their lower labour market participation and/or worse wage conditions.
Furthermore, these results also show the social response to this situation: delayed emancipation or even returning to the parental home, which is mainly reflected in the increase in the weight of imputed rent. As highlights, this pattern is characteristic of southern European welfare regimes, where weak public support for young people results in an extended dependency on family networks. The prolonged cohabitation with parents functions as an informal safety net that compensates for insufficient wages and precarious employment. But the dependence on family support highlights the structural weaknesses of public institutions, which have historically given priority to older generations in terms of social services.
The results on the evolution of income distribution and poverty by age groups in Galicia show that young people have experienced a decline in their income levels. Consequently, income inequality has increased, as has the percentage of young people living in poverty.
To complete the analysis, it is necessary to examine the changes in the determinants of income distribution and poverty across the different age groups. Figures 10, 11 and 12 illustrate the changes in the relevance of determinants in explaining income distribution among young individuals, middle-aged individuals, and older adults, while Figure 13, 14 and 15 do the same for explaining poverty incidence.
The results presented in Figures 10, 11 and 12 highlight several interesting trends regarding the evolution of income decile determinants for individuals based on their age profile, with the following four standing out:
In the young group, the most relevant explanatory factors are related to employment status, gender, and educational attainment. Among the employment-related variables, being employed (emp1) or studying (emp4) are highly significant, although their importance decreases over the period. Within the educational variables, having higher education (edu5) stands out as the most relevant factor, despite its declining relevance throughout the period. On a second level of importance are factors such as the type of municipality, occupation, and the tenure type of the primary residence. Specifically, owning a home (tenure1) emerges as a relevant factor, and among the occupational variables, working in jobs related to hospitality and retail (emp6) is also important. In terms of territorial scope, the three variables that take into account the population density of the municipality are of similar relevance. This shows that the concentration of young people in the lower income strata is not exclusively an urban phenomenon.
Among middle-aged individuals, three variables stand out in explaining their distribution across income deciles: gender, being employed (emp1), and engaging in household duties (occu6). Notably, the relevance of household duties has increased over the period. A second group of significant explanatory factors includes variables related to education, type of municipality, housing tenure, and type of occupation. It is noteworthy that educational attainment, type of municipality, and, especially, housing tenure have gained relevance in explaining individuals’ income decile over time, whereas gender, employment status, and occupation-related variables have lost importance. This shift highlights the growing role of access to housing as a determinant of well-being for middle-aged individuals. It can also be observed that within this group, the type of municipality is a determining factor, especially the fact of belonging to smaller municipalities.
In the group of older adults, the most prominent explanatory factors relate to gender, educational attainment, and employment status. Specifically, having higher education (edu5) emerges as a highly significant factor, with its importance increasing considerably over the period. However, having primary or secondary education also appears to be relevant. Regarding employment status, being retired (emp4) and engaging in household duties (emp6) are the most significant factors, both of which have grown in importance over the period. Within this group, the type of municipality is a determining factor, with the fact of living in a sparsely populated municipality standing out above the rest.
Figures 13, 14 and 15 show the evolution of the relevance of explanatory factors in poverty incidence across different age groups. Three main conclusions can be drawn from the analysis:
Within the group of young individuals, two explanatory factors stand out due to their high relevance: being employed (emp1) and studying (emp3). However, it is noteworthy that both of these factors have lost considerable relevance over the period. Among middle-aged individuals, two factors also stand out: being employed (emp1) and engaging in household duties (emp6). For older adults, the key explanatory factors are being retired (emp4) and engaging in household duties (emp6), increasing its relevance during the analysed period.
Regarding the variation in the relevance of productive factors, it should be noted that among middle-aged individuals, being employed (emp1) has significantly increased in relevance. The same trend is observed for being unemployed (emp2). In contrast, engaging in household duties (emp6) has lost relevance over the period.
The results obtained show that, over the past few decades, particularly since the Great Recession, the well-being of young people in Galicia (those under the age of 35) has deteriorated. Specifically, the findings reveal that young individuals are increasingly concentrated in the lower income deciles, while middle-aged adults (aged 35 to 64) and older individuals (over 65) have experienced upward social mobility. This shift is also reflected in a sharp rise in the incidence of poverty among the youngest age groups. Moreover, an analysis of the determinants of poverty and income among young people reveals that most of these factors have lost explanatory power, highlighting the widespread and persistent nature of poverty among youth throughout the period.
Underlying the deterioration in the well-being of young people over recent decades are, among other factors, a decline in their labour market participation and an increase in job insecurity. Furthermore, as evidenced by the results obtained, public policies have done significantly less to offset the loss of well-being among young people compared to the other two age groups.
These results show that the crisis has disproportionately affected young people, concentrating its negative effects on this group. As a consequence, several related issues arise that merit further discussion. The fact that income losses are concentrated among the young, with long-lasting effects over time, highlights the need to introduce new income replacement policies targeted at younger individuals and/or to expand existing ones. In response to this need, the introduction of a basic income scheme specifically aimed at young people could be a promising alternative, enabling them to cope with greater income volatility. This proposal aligns with the ideas of authors such as , who advocates for a substantial one-off payment upon reaching adulthood, which he refers to as a "universal inheritance". Such a policy, however, runs the risk of undermining horizontal equity, as it provides the same payment regardless of the individual's socio-economic status. For this reason, it seems pertinent to also consider support measures targeted at individuals experiencing poverty or precarious living conditions.
As a result of the decline in young people's well-being, further exacerbated by rising rental prices, the generational gap has widened significantly in Galicia. It is therefore essential to place youth well-being at the centre of the political agenda. This issue is all the more pressing given that young people constitute a demographic group with relatively low political weight and participation due to both demographic and social factors. This may help explain the limited attention paid to young people by the main political parties and may partially account for the rise and consolidation of far-right populism among them.
Thus, the reasons behind young people's discontent, are not only rooted in the sharp deterioration of their living conditions, as demonstrated in this article, but also in the neglect they have experienced from public policy. It is therefore essential to move towards a social pact that addresses the living conditions of young people, so that they may be compensated, at least in part, for bearing the brunt of economic crises without reaping the benefits of subsequent recoveries. Without such a pact and effective policies, Galician society —and Spanish society more broadly— may face a scenario of generational conflict that undermines social cohesion between age groups and fuels the growth of political movements that exacerbate such divisions, despite offering no real solutions to the challenges at hand.
Another aspect that emerges from the analysis is the significant role of family support during economic crises, which is generally provided from older to younger generations. While this is undoubtedly true and relevant, it is worth questioning to what extent it is effective and desirable for the main source of support for young people to come from families rather than directly from the public sector. The main concern underlying this question lies in the unequal capacity of households to provide support and the potential dependency relationships it may foster. In a society with the demographic situation and trends currently observed in Galicia, making young people’s minimum well-being dependent on their families does not appear to be the most appropriate strategy for addressing the region’s natural demographic decline. Therefore, it seems more appropriate to direct policies explicitly towards young people—or, at the very least, towards those who lack sufficiently strong family support networks to sustain a minimum and stable standard of living.
This research analyses the changes in income distribution and poverty incidence in Galicia between 2007 and 2021 focusing on the generational disparities. The findings of this study show that young people have been the most affected by adverse economic dynamics over the period analysed, with the impact of the pandemic being particularly striking. Specifically, the results demonstrate that those under 35 have experienced a reduction in their income levels, an increase in income inequality, and higher levels of poverty incidence. In contrast, middle-aged adults (aged 35 to 64) and older adults (over 65) have seen an increase in their income levels during the period and experience lower levels of poverty by the end of the period.
These different dynamics between young people and other adults are explained in part by the higher dependence of young people on market income, who face higher rates of unemployment, job insecurity and lower wages, as well as limited access to stable jobs. On the other hand, middle-aged and older adults have benefited from greater coverage of social benefits, such as pensions and public transfers, as well as greater employment stability, particularly in sectors with less exposure to temporary and precarious employment. This imbalance also highlights that current public policies are limited when it comes to protecting young people from the effects of economic crises.
The analysis underscores that the income sources and the income distribution and poverty determinants differ markedly across age groups. For young people, the decline in wages and limited access to stable employment have contributed to their welfare reduction. Middle-aged individuals, while generally are better positioned, also face challenges, particularly concerning housing access and employment stability. Older adults, benefiting from a robust social safety net, remain relatively insulated from poverty but still experience inequalities in income distribution.
These results point to the urgent need for targeted economic and social policies. Addressing youth poverty and income inequality requires not only improving labour market conditions but also implementing youth-focused social benefits and housing support programs. At the same time, policies must ensure the continued protection of older adults and support the middle-aged population in maintaining economic stability.
Educational attainment has behaved differently across age groups. It has increased in importance as an explanatory factor among older adults, where higher education has become a key determinant of higher earnings. Among young people, however, its importance has declined slightly, highlighting the limitations of the labour market to reward educational attainment in this group. This could be explained by the processes of educational expansion or a lower value of qualifications for generating educational returns.
Housing has been identified as a key factor in income distribution and poverty incidence, especially for young and middle-aged adults. The results show an increase in the role of tenure in explaining income inequalities, especially for middle-aged adults, while for young people there is an increasing dependence on imputed rents. This phenomenon reflects the difficulty for young people to access home ownership or rental housing, forcing them to delay their emancipation or even return to the parental home, as had already pointed out for southern European countries. In this context, it is essential to move towards a new social contract that recognises and addresses the deterioration in young people's living conditions. This should be reflected in ambitious, sustained public policies that promote the well-being of young people, reduce their reliance on family support and strengthen social cohesion in the face of increasingly challenging demographic trends.
By identifying the evolving determinants of income distribution and poverty, this study provides a foundation for policymakers to design interventions that promote equity and social cohesion across generations. Future research could further investigate the long-term effects of these disparities and the impact of policy interventions on reducing generational inequality.
This research has been supported by the ICEDE research group, to which the authors belong, Galician Competitive Research Group ED431C 2022/15 financed by Xunta de Galicia and project “REVALEC” REFERENCE PID2022-141162NB-I00 Financed by MCIN/ AEI / 10.13039/501100011033 / EFRD, EU.
Also, José Manuel Amoedo the support received from the Xunta de Galicia 2021 predoctoral training support programme (ayudas de apoyo a la etapa de formación predoctoral) with reference number ED481A 2021/084. Hugo Campos-Romero acknowledges the support received from the Xunta de Galicia 2024 postdoctoral training support programme (ayudas de apoyo a la etapa de formación posdoctoral), co-funded by the Consellería de Cultura, Educación, Formación Profesional y Universidades and the Agencia Gallega de Innovación (reference number ED481B_048), as well as the funding received by the European Union EACEA for the project “CEBCAT”, reference 101179061, financed by the Erasmus+ Programme.
1
Amoedo, J. M., & Sánchez-Carreira, M. del C. (2023). The initial effects of Covid-19 on income inequalities in Galicia and their distribution in society. Revista Galega de Economía, 32(1), 1-23. https://doi.org/10.15304/rge.32.1.8551
2
4
Bentolila, S., Felgueroso, F., Jansen, M. & Jimeno, J. F. (2022). Lost in recessions: youth employment and earnings in Spain. SERIEs, 13, 11-49. https://doi.org/10.1007/s13209-021-00244-6
5
Blanco, O., & Julián, D. (2019). A typology of precarious employment for Chile: Precariousness as a cross-class phenomenon. CEPAL Review, 129, 91-128. https://doi.org/10.18356/8f5ac9c8-es
6
Boertien, D., & López-Gay, A. (2023). The polarization of real estate ownership and increasing wealth inequality in Spain. European Sociological Review, 39(4), 615-629. https://doi.org/10.1093/esr/jcac072
7
Breiman, L. (1996). Bagging predictors. Machine learning, 24, 123-140. https://doi.org/10.1007/BF00058655
8
Breiman, L. (2001). Random Forests. Machine Learning, 45, 5–32. https://doi.org/10.1023/A:1010933404324
9
10
11
Chauvel, L. (2010). The long-term destabilization of youth, scarring effects, and the future of the welfare regime in post-trente glorieuses France. French Politics, Culture & Society, 28(3), 74-96. https://doi.org/10.3167/fpcs.2010.280305
12
Chybalski, F., & Marcinkiewicz, E. (2022). A multidimensional approach to intergenerational balance measurement: a cross-sectional study for European countries. Social Justice Research, 35, 206-241. https://doi.org/10.1007/s11211-022-00392-5
13
Corbelle, F., & Troitiño, A. (2013). Inequality and Poverty in Galicia between 2007 and 2010. How is the Distribution of Crisis Effects? Revista galega de economía, 22(2), 167-200. https://doi.org/10.15304/RGE.22.EXTRA.1406
14
15
Díez, J. M., & Pardo, A. (2020). Despoblación, envejecimiento y políticas sociales en Castilla y León. Revista Galega de Economía, 29(2), 1-18. https://doi.org/10.15304/rge.29.2.6959
16
17
18
Escalonilla, M., Cueto, B., & Pérez-Villadóniga, M. J. (2022). Is the millennial generation left behind? Inter-cohort labour income inequality in a context of economic shock. Social Indicators Research, 164, 285-321. https://doi.org/10.1007/s11205-022-02958-x
19
Eurostat (2024). Living conditions in Europe - poverty and social exclusion. Statistics Explained. https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Living_conditions_in_Europe_-_poverty_and_social_exclusion
20
Fernández-Leiceaga, X. (2024). Demographic regression and local finances: the case of the Galician municipalities 2001-2019. Revista Galega de Economía, 33(3), 9601. https://doi.org/10.15304/rge.33.3.9601
21
de la Fuente, A. (2021). The economic consequences of Covid in Spain and how to deal with them. Applied Economic Analysis, 29(85), 90-104. https://doi.org/10.1108/AEA-11-2020-0158
22
23
Grömping, U. (2015). Variable importance in regression models. Wiley interdisciplinary reviews. Computational statistics, 7(2), 137-152. https://doi.org/10.1002/wics.1346
24
Hess, M., Nauman, E., & Steinkopf, L. (2017). Population ageing, the intergenerational conflict, and active ageing policies–a multilevel study of 27 European countries. Journal of Population Ageing, 10, 11-23. https://doi.org/10.1007/s12062-016-9161-3
25
Carrascal, A., & Fernández, M. (2015). Tourism and income distribution: Evidence from a developed regional economy. Tourism Management, 48, 11-20. https://doi.org/10.1016/j.tourman.2014.10.016
26
27
28
Liu, M., Hu, S., Ge, Y., Heuvelink, G. B., Ren, Z., & Huang, X. (2021). Using multiple linear regression and random forests to identify spatial poverty determinants in rural China. Spatial Statistics, 42, 100461. https://doi.org/10.1016/j.spasta.2020.100461
29
Merino-Llorente, M. C., & Arechavala, M. N. (2020). European part-time workers’ health and well-being in times of crisis. The case of female part-timers. Hacienda Pública Española, 235(4), 61-86. https://dx.doi.org/10.7866/HPE-RPE.20.4.4
30
Módenes, J. A. (2022). Inestabilidad y problemas de acceso a la vivienda, una realidad cada vez más extendida: Evolución reciente de las desigualdades demográficas y sociales en el acceso y la estabilidad residenciales. In Ayala, L. (Ed.), Desigualdad y pacto social (pp. 59-76). El Observatorio Social, Fundación la Caixa.
31
Moreno, A. (2012). The transition to adulthood in Spain in a comparative perspective: The incidence of structural factors. Young, 20(1), 19-48. https://doi.org/10.1177/110330881102000102
32
OECD [Organisation for Economic Co-operation and Development] (2022). Delivering for youth: How governments can put young people at the centre of the recovery. OECD Policy Responses to Coronavirus (COVID-19). OECD Publishing. https://doi.org/10.1787/92c9d060-en
33
Pena-Boquete, Y., & Dios-Murcia, I. (2021). Factors behind the employment loss in Galicia: Great Recession of 2008 vs. the first wave of the COVID-19 pandemic. Revista Galega de Economía, 30(1), 1-18. https://doi.org/10.15304/rge.30.1.7451
36
Rice, J. M., Temple, J. B. & McDonald, P. F. (2021). Intergenerational inequality and the intergenerational state. Journal of Population Research, 38(4), 367-399. https://doi.org/10.1007/s12546-021-09273-1
37
38
Sánchez, M. C., Sánchez, P., Cruz, M. M., & Sánchez, F. J. (2013). Precariedade laboral-Pobreza en Galicia en tempos de crise. Revista Galega de Economía, 22, 201-224. https://doi.org/10.15304/rge.22.Extra.1407
39
40
Sohnesen, T. P., & Stender, N. (2017). Is random forest a superior methodology for predicting poverty? An empirical assessment. Poverty & Public Policy, 9(1), 118-133. https://doi.org/10.1002/pop4.169
41
Soriano, I., & Gainza, X. (2025). La Desigualdad de Riqueza en Perspectiva Generacional: Evidencia de la Encuesta Financiera de las Familias (2002-2020). Revista Española de Investigaciones Sociológicas, (189), 149–166. https://reis.cis.es/index.php/reis/article/view/60
42
43
Úbeda, M., Cabasés, M. À., Sabaté, M., & Strecker, T. (2020). The Deterioration of the Spanish Youth Labour Market (1985–2015): An Interdisciplinary Case Study. YOUNG, 28(5), 544-563. https://doi.org/10.1177/1103308820914838
44
Van de Velde, C. (2022). ¿Una brecha generacional global?: Juventud y relaciones intergeneracionales en el siglo xxi. In L. Sepúlveda & A. Moreno (Eds.), Transiciones educativo-laborales de jóvenes en tiempos de incertidumbre (pp. 53–78). Ediciones Universidad Alberto Hurtado. https://doi.org/10.2307/j.ctv3596xhc
45
Vázquez-Taín, M. Á. (2020). Demografía y sostenibilidad del sistema público de pensiones. Revista Galega de Economía, 29(2), 1-18. https://doi.org/10.15304/rge.29.2.6917
Source: The authors from
[1] By way of clarification, a brief description of the following terms is given below: vulnerability, precariousness, relative poverty and absolute poverty. Vulnerability is defined as a situation that generates an increased likelihood of falling into poverty or precariousness in the face of changes in the environment. Precariousness refers to the existence of instability in fundamental aspects such as employment, housing or access to basic services. The difference between relative poverty refers to how the two are measured. On the one hand, relative poverty measures the situation of an individual in comparison with the rest by establishing a moving threshold (commonly used for income is 60% of the median income). On the other hand, absolute poverty measures the individual's situation by setting a fixed threshold (e.g. having an income below a certain amount).
[2] It measures the percentage of people at risk of poverty or social exclusion, understood as the situation of those who meet at least one of three criteria: having a disposable income below 60% of the national median (at risk of poverty), living in conditions of severe material and social deprivation (lacking at least 7 out of 13 essential elements), or residing in households with low employment intensity (adults worked less than 20% of their labour potential in the year). The difference between relative poverty refers to how both are measured. On the one hand, relative poverty measures the individual's situation in comparison with the rest by setting a moving threshold (commonly used for income at 60% of median income). On the other hand, absolute poverty measures the individual's situation by setting a fixed threshold (e.g. having an income below a certain amount).
