The Feminisation of Poverty in European Union Countries—Myth or Reality?

Abstract

The feminisation of poverty is a complex, multidimensional phenomenon related to gender inequality in various aspects of life. Women are disproportionately affected by the gender pay gap, unequal intra-household resource distribution, unpaid domestic work, caregiving responsibilities, single motherhood, employment or educational barriers, violence, gender discrimination, and period poverty. Combating poverty and inequality are among the main goals of the 2030 Agenda for Sustainable Development adopted by the United Nations General Assembly on 25 September 2015, and their great importance is further highlighted in the preamble of the resolution outlining the plan to transform our world by 2030. This study uses SDG indicators from the Eurostat database to assess the feminisation of poverty in the EU-27 member states in 2020 based on selected diagnostic characteristics mainly related to SDG 1 and SDG 5; they are also related to other goals due to the cross-cutting nature of the topic. The characteristics were transformed to reflect gender gaps and afterwards unitised with a veto threshold indicating gender balance. These were then used to calculate a synthetic taxonomic measure, allowing for linear ordering and classification of countries based on the feminisation of poverty levels. The study confirmed significant feminisation of poverty in the EU-27, with a few aspects showing masculinisation. High feminisation of poverty was observed in both emerging and advanced economies.

1. Introduction

The lack of a universally accepted definition of poverty has led to its examination from diverse theoretical perspectives. Codes A.L. [1] outlines several core approaches, including subsistence (material poverty), basic necessities, and relational deprivation. When delineating poverty in terms of subsistence, it is fundamentally tied to the essential requirements for human survival, encompassing basic nutritional and physical needs. The basic needs perspective includes clothing, education, transportation, and other necessities beyond nutritional and physical demands. Relational deprivation pertains to an individual’s social sphere and their interactions with the environment. Sen and Grusky et al. [2,3] define poverty not merely as a lack of income but as the inability to lead a meaningful life due to a shortage of economic resources, thus providing a multifaceted perspective that includes social and political dimensions.

Similarly, the concept of sustainability should be approached in a multifaceted way, taking into account its three dimensions: environmental, economic, and social [4]. At the same time, the implementation of this concept should be supported by analyses so that it becomes pragmatic sustainability in real life [5,6]. The Sustainable Development Goals (SDGs) of Agenda 2030 [7,8], emphasise tackling gender poverty and promoting gender equality [9]. Achieving SDG 1, “No poverty”, means addressing the root causes of gender-based poverty to end poverty for all, including women and girls. The implementation of SDG 5, “Gender equality”, is to ensure the achievement of gender equality and the empowerment of all women and girls, focusing on equal access to education, economic opportunities, and participation in decision-making. Due to the multifaceted nature of sustainability, references to both also appear in other Sustainable Development Goals. Agenda 2030 recognises that gender equality and women’s empowerment are integral to all SDGs. Efforts to combat gender poverty include measures such as equal pay, quality education and healthcare access, women’s empowerment in decision-making, and eliminating gender-based violence.

Both women and men experience gender-based inequalities, but it is mostly women who face poverty because of it. The term “feminisation of poverty”, which refers to the increased representation of women among the impoverished or economically disadvantaged, was originally coined by Diana Pearce [10]. In her 1978 work, “The Feminization of Poverty: Women, Work, and Welfare”, Pearce highlighted that in the United States, two-thirds of those living in poverty were women aged 16 and above [11]. This phenomenon primarily relates to the overrepresentation of women and children in lower socioeconomic status groups compared to men of the same socioeconomic status. The causes of feminisation of poverty include factors such as family and household structures, employment opportunities, violence, access to education, the impact of climate change, economic disparities affecting women (femonomics), and health disparities. Persistent traditional stereotypes regarding women continue to be ingrained in many cultures, limiting income opportunities and community participation for many women. Today, the concept of the feminisation of poverty has evolved into a broader and more complex idea. It must be examined in the context of the social inequalities that women experience. Women not only earn lower wages and pensions than men but also face higher levels of unemployment (being not employed but available for work and actively seeking work), including long-term unemployment (for 12 months or more), the challenges of single motherhood, and the burdens of household responsibilities [12].

Among the most common forms of female poverty, literature studies include:

• Decision-making power—women often experience decision-making poverty, meaning that they do not have an equal say in decisions about their lives, whether family, social, or professional [13,14,15];
• Disparate income (gender pay gap)—income inequality, often referred to as the gender pay gap, signifies that women receive lower earnings than men in identical or similarly qualified positions. This disparity in income stands as a primary factor contributing to the heightened risk of poverty among women. Insufficient income hinders women from acquiring assets and translating their financial resources into social and economic standing. Moreover, higher income not only provides increased opportunities for skill development but also leads to higher earnings [16,17,18,19];
• Energy poverty—means lack of access to essential energy, which can affect women’s quality of life and health, especially in low-income countries and in households where the woman is the head of the family [20,21,22];
• Lack of assets—a primary element exacerbating poverty among women is their restricted ability, potential, and authority to reach and manage productive resources, which encompass land, employment, human capital assets, such as education and health, and social capital assets, such as engagement at various levels [23,24], legal entitlements, and safeguards [25];
• Time poverty—time is a component integrated into the concept of poverty because it represents a fundamental resource that is frequently distributed unevenly among individuals, particularly when other resources are inadequate. Women unquestionably experience a greater scarcity of time compared to men [26,27]. Research reveals that in addition to their paid employment, women are also heavily engaged in reproductive and unpaid tasks (domestic work) [28,29,30]. The distribution of time between men and women within the household and the broader economy is a significant gender-related issue in the ongoing discussion on time scarcity. Time poverty can be understood in terms of the insufficient availability of time for rest and sleep. The more time dedicated to paid or unpaid work, the less time remains for other activities like relaxation and enjoyment;
• Capability deprivations—lack of access to opportunities, for example, educational and vocational, can affect women’s opportunities for development and their ability to improve their social and economic status [31,32,33];
• Deprivation of health outcomes—lack of access to health care, health education, and adequate living conditions can result in poverty in health outcomes and increase women’s risk of health problems, particularly in underdeveloped countries [34];
• Period (menstrual) poverty—limited or no access to menstrual products, sanitation and hygiene facilities, and education and awareness to manage menstrual health [35,36,37]. This can affect women’s health and dignity, and lower their involvement in work or education [38];
• Social and cultural exclusions—refers to a situation where women may be excluded from social and cultural participation due to stereotyping, discrimination, and gender inequality. This can affect their access to resources and opportunities [39].

An analysis of work from recent years shows that the COVID-19 pandemic has significantly exacerbated the feminisation of poverty, in terms of most of the categories mentioned earlier. The economic downturn caused by the pandemic disproportionately affected women, particularly in low-income countries and vulnerable sectors like the service industry, where women are overrepresented [40]. Additionally, the pandemic intensified social inequalities, leading to increased job losses among women, a rise in unpaid domestic work, and greater barriers to accessing healthcare and education [41]. These factors have collectively deepened the gender disparities in poverty, reinforcing the economic and social vulnerabilities of women worldwide [42].

The feminisation of poverty is measured by analysing gender differences in the context of poverty and economic inequality. There are several indicators and approaches used to measure the feminisation of poverty. Poverty based on income compares the incomes of women and men, paying attention to the gender earnings gap. If women have significantly lower incomes than men, this can be considered a manifestation of the feminisation of poverty [43,44]. Using official poverty definitions, poverty rates for women and men can be calculated and compared. If poverty rates are higher among women, then we can talk about the feminisation of poverty [45,46,47]. The Gini Index measures income inequality within a given population. A higher Gini coefficient means greater income inequality. Comparing the Gini index for men and women can reveal income inequality between the genders [48,49,50,51]. Analysing employment and unemployment rates between genders can provide information about differences in access to work and labour market conditions [52,53,54]. Education is an important factor in combating poverty. Analysing educational indicators such as educational attainment and access to education can help understand the extent to which women have access to tools to avoid poverty [55,56]. Analysis of health indicators and access to healthcare highlights possible health inequalities that may impact women’s risk of poverty [57,58,59].

Based on the above indicators, the feminisation of poverty is measured primarily by three international indicators that provide comprehensive analysis and data on gender inequality in the context of poverty and the role of women in society. The Gender Development Index (GDI) is an indicator developed by the United Nations Development Program (UNDP) that combines the traditional Human Development Index (HDI) with a focus on gender gaps. The GDI takes into account not only economic indicators such as per capita income and educational attainment but also gender disparities in these areas. The GDI helps understand how women and men participate in social and economic development processes [60,61,62,63]. The Gender Empowerment Measure (GEM) is another indicator developed by UNDP. It measures women’s level of autonomy and influence in society. The GEM assesses gender inequalities in women’s participation in public life, politics, and the economy. This helps to understand the extent to which women have the opportunity to participate and make decisions in these areas [60,64,65]. The Human Poverty Index (HPI) is a United Nations index that takes into account various aspects of poverty such as health, education, and standard of living. In the context of the feminisation of poverty, the HPI also examines gender inequalities in these areas. This allows us to understand how gender disparities affect poverty levels in a given population [66,67,68].

A literature review on the feminisation of poverty in European Union (EU) countries reveals a complex and multifaceted issue. The term “feminisation of poverty” refers to the disproportionate representation of women among the world’s poor, often resulting from gender-based inequalities and discrimination. In the context of EU countries, this phenomenon manifests itself in various ways, and research has explored these dynamics extensively. This literature review provides an overview of key findings and examples from EU member states to illustrate the feminisation of poverty. Economic disparities and the gender pay gap are persistent problems in all EU member states. Women tend to earn less than men, resulting in lower lifetime earnings and reduced economic security. Research by the European Institute for Gender Equality (EIGE) highlights the gender pay gap as a key contributor to the feminisation of poverty in the EU [69]. According to Eurostat, for the economy as a whole, in 2022, women’s gross hourly earnings were on average 12.7% below those of men in the European Union and 13.2 % in the euro area [70]. The problem of the Gender Wage Gap in EU countries has been addressed by many authors [71,72,73,74]. Who, in addition to analysing the gender pay situation in individual EU countries, also study the impact of eliminating the gender pay gap on income and poverty [19]. Women in the EU are more likely to engage in part-time and precarious employment, which often leads to financial vulnerability. Studies have shown that part-time work can limit access to social security benefits and career advancement, thereby increasing the risk of poverty [75,76,77,78]. Single-parent households, predominantly led by women, face higher poverty rates. The lack of support and affordable childcare services can make it difficult for single mothers to escape poverty. A study around the EU (and OECD countries) explores the experiences of single parents, especially single mothers living in poverty, and their challenges in accessing social services [79,80,81,82,83,84]. Women often bear the brunt of unpaid care work, such as taking care of children, the elderly, or family members with disabilities. This can limit their participation in the labour market and their ability to earn a living. An analysis of the unequal distribution of care work in EU countries can be found in [85,86,87,88]. Intersectionality plays a crucial role in understanding the feminisation of poverty. Immigrant and minority women may face compounded forms of discrimination that contribute to their higher poverty rates [89,90,91]. The effectiveness of social welfare policies in addressing the feminisation of poverty varies among EU member states. Research on the impact of welfare policies on gender equality in different EU countries is discussed by Sánchez-Lopez and de Paz Báñez [92], McDevitt [93], or Ostner and Lewis [94]. Within the analyses of the Social Welfare Policies of the EU countries, the problem of poverty among the elderly, including women, is also frequently raised [95,96,97,98].

This paper adopts the concept of the feminisation of poverty being a multidimensional phenomenon that is related to gender inequality occurring in various aspects of life. The purpose of this study was to examine the presence of the feminisation of poverty in European Union member states and to compare its level in these countries based on selected SDG indicators. Variables from the Eurostat database that directly represent poverty, as well as those that can influence it, were selected for the study. To measure the level of feminisation of poverty, the use of transformed gender gaps was proposed. The specific objectives of the study were to rank countries according to the level of feminisation of poverty and to create classes of countries similar in terms of their level of feminisation of poverty. The empirical study carried out allowed the following questions to be answered: (1) Is there a feminisation of poverty in EU member states? (2) By which diagnostic characteristics does the feminisation of poverty occur? (3) Are there diagnostic characteristics that contribute to the masculinisation of poverty or because of which there is a gender balance in terms of poverty? (5) Does the level of feminisation of poverty vary across EU member states? (6) Which countries have the highest and the lowest levels of feminisation of poverty? (7) Which countries are similar with regard to the level of feminisation of poverty?

2. Materials and Methods

The empirical study covered the 27 member states of the European Union (EU-27). It used statistics showing the selected SDG indicators for 2020 by sex from the database of the statistical office of the European Union, Eurostat. The selection of primary data took into account both variables representing the level of poverty directly (e.g., the percentage of people at risk of income poverty or experiencing severe material deprivation) and those that may contribute to or counteract it (e.g., access to education, employment and percentage of inactive people due to caring responsibilities). Indicators on health or access to water and sanitation that relate to period poverty were also included in the study [99,100]. The primary variables collected were related to the Agenda 2030 Goals: SDG 1 “No poverty”, SDG 3 “Good health and well-being”, SDG 4 “Quality education”, SDG 5 “Gender equality”, SDG 6 “Clean water and sanitation”, SDG 8 “Decent work and economic growth”; they were also related to other goals due to the cross-cutting nature of the topic. In order to select variables with the highest possible diagnostic values, the primary data set was subjected to selection according to statistical criteria [101]. To eliminate variables with low variability and those highly correlated with others, cut-off values of Pearson’s linear correlation coefficient and classical coefficient of variation of 0.7 and 10%, respectively, were adopted [102]. Due to the sufficiently high coefficient of variation values, Pearson’s coefficient proved to be the main statistical criterion for selection. Two primary variables, SDG 1.10 “Persons at risk of poverty or social exclusion” and SDG 11.11 “Severe housing deprivation rate”, highly correlated with several other variables and were finally excluded from the set of collected sustainability indicators. However, this did not result in a significant loss of information, as the removed variable SDG 1.10 is a combination of indicators SDG 1.20, SDG 1.31, and SDG 1.40, and they remained in the final set.

Finally, 15 diagnostic features were selected, the values of which for the study objects (EU-27 member states) by sex were collected in matrices:

X = [x_ij]_27×15,

(1)

Y = [y_ij]_27×15,

(2)

where:

• i—object number (i = 1, 2, …, 27),
• j—diagnostic feature number (j = 1, 2, …, 15),
• X—matrix of values of all diagnostic features for women and for all objects,
• x_ij—the value of the j-th diagnostic feature for women and for the i-th object,
• Y—matrix of values of all diagnostic features for men and for all objects,
• y_ij—the value of the j-th diagnostic feature for men and for the i-th object.

The columns of the X and Y matrices contain the values of the diagnostic features for all objects by sex, i.e., X_j and Y_j represent the j-th diagnostic feature for women and men respectively.

Table 1. Selected descriptive characteristics of diagnostic features by sex.

j	Diagnostic Feature (in %)	Xj					Yj
		Mean	Median	Min	Max	Skewness	Mean	Median	Min	Max	Skewness
1	SDG 1.20 People at risk of income poverty after social transfers	17.1	16.7	11.7	25.8	0.54	15.3	15.0	7.2	22.1	0.10
2	SDG 1.31 Severe material and social deprivation rate	6.7	4.7	1.4	26.5	2.17	6.1	4.3	1.5	24.0	2.20
3	SDG 1.40 People living in households with very low work intensity	7.6	7.6	3.6	13.0	0.36	7.2	7.2	3.9	11.9	0.30
4	SDG 1.41 In work at-risk-of-poverty rate	6.9	7.3	3.0	11.4	0.08	8.6	8.1	3.3	18.1	0.78
5	SDG 1.50 Housing cost overburden rate	7.6	6.3	1.8	34.2	3.21	6.9	5.2	2.0	32.3	3.51
6	SDG 3.60 Self-reported unmet need for medical examination and care	2.7	1.7	0.0	14.6	2.35	1.9	1.4	0.0	11.1	2.70
7	SDG 4.10 Early leavers from education and training	7.0	6.3	2.0	16.6	1.03	10.3	10.0	2.4	20.2	0.27
8	SDG 4.20 Tertiary educational attainment	35.0	37.7	17.5	46.7	−0.32	27.3	26.6	15.0	40.2	−0.06
9	SDG 4.31 Participation in early childhood education	90.4	92.0	71.5	100	−1.05	90.2	91.6	71.0	100.0	−1.03
10	SDG 4.60 Adult participation in learning	11.3	8.7	1.0	35.5	1.35	8.8	7.4	1.0	23.0	0.96
11	SDG 5.40 Inactive population due to caring responsibilities	32.5	30.4	7.5	66.6	0.32	12.3	8.1	1.5	41.4	1.44
12	SDG 6.10 Population having neither a bath, nor a shower, nor indoor flushing toilet in their household	1.9	0.4	0.0	21.3	3.76	2.0	0.4	0.0	21.1	3.51
13	SDG 8.20 Young people neither in employment nor in education and training	16.9	17.1	7.7	29.3	0.34	11.1	10.6	5.0	21.0	0.79
14	SDG 8.30 Employment rate	68.5	70.6	48.7	77.4	−1.24	78.3	78.2	68.1	87.2	−0.11
15	SDG 8.40 Long-term unemployment rate	2.4	1.8	0.6	13.5	3.46	2.2	1.9	0.5	8.2	2.30

summarises selected descriptive characteristics of diagnostic features by sex. Most of the diagnostic features (for j = 1, 2, …, 7, 11, 12, 13, 15) are poverty stimulants, which are positively correlated with the level of the complex phenomenon of poverty. The remaining four variables (for j = 8, 9, 10, 14) are destimulants of poverty, i.e., are negatively correlated with the level of poverty. Most of the diagnostic features are characterised by right-handed asymmetry, while left-handed asymmetric are the variables X₉, X₁₄, and Y₉, which are destimulants of poverty. Comparing the descriptive parameters of the features for men and women, it can be seen that the median of most of the diagnostic features has larger values for women. The largest difference of 22.3 pp. is for the stimulant representing inactive population due to caring responsibilities (j = 11), followed by 11.1 pp. for the destimulant of tertiary educational attainment (j = 8). A smaller median value for women applies to only four features (numbered j = 4, 7, 14, 15), the largest difference (7.6 pp.), to women’s disadvantage, is for the destimulant representing employment rate (j = 14). Similar relationships apply to the other parameters of the diagnostic features for men and women—the mean and the extremes.

In the empirical research carried out, the pattern-free method of multivariate comparative analysis was used to assess the level of feminisation of poverty in the EU-27 member states. The research procedure followed six steps.

Step 1. Determination of the matrix:

V = [v_ij]_27×15,

(3)

with elements representing gender gaps (cf. [103]):

$${\mathrm{v}}_{\mathrm{i}\mathrm{j}}=\left\{\begin{array}{cc}{\mathrm{x}}_{\mathrm{i}\mathrm{j}}-{\mathrm{y}}_{\mathrm{i}\mathrm{j}}& \mathrm{f}\mathrm{o}\mathrm{r} \mathrm{s}\mathrm{t}\mathrm{i}\mathrm{m}\mathrm{u}\mathrm{l}\mathrm{a}\mathrm{n}\mathrm{t}\mathrm{s}\\ {\mathrm{y}}_{\mathrm{i}\mathrm{j}}-{\mathrm{x}}_{\mathrm{i}\mathrm{j}}& \hspace{1em}\mathrm{f}\mathrm{o}\mathrm{r} \mathrm{d}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{i}\mathrm{m}\mathrm{u}\mathrm{l}\mathrm{a}\mathrm{n}\mathrm{t}\mathrm{s}\end{array}\right.$$

(4)

Variable V_j being the j-th column of the V matrix represents gender gaps due to the j-th diagnostic characteristic for all objects (EU-27 member states). Its values indicate the direction of gender imbalance (positive values indicate female disadvantage and negative values indicate male disadvantage) or indicate gender balance (values equal to 0).

Step 2. Transformation of matrix V to form:

Z = [z_ij]_27×15,

(5)

where

$${\mathrm{z}}_{\mathrm{i}\mathrm{j}}=\left\{\begin{array}{cc}0& \mathrm{f}\mathrm{o}\mathrm{r} {\mathrm{v}}_{\mathrm{i}\mathrm{j}}\le 0\\ {\mathrm{v}}_{\mathrm{i}\mathrm{j}}& \mathrm{f}\mathrm{o}\mathrm{r} {\mathrm{v}}_{\mathrm{i}\mathrm{j}}>0 \end{array}\right.,$$

(6)

Rejecting negative values from the matrix V transforms its columns into the form of the stimulants of feminisation of poverty (Z_j).

Step 3. Normalisation of the stimulants of feminisation of poverty:

$${\mathrm{z}}_{\mathrm{i}\mathrm{j}}^{\mathrm{*}}={\displaystyle \frac{{\mathrm{z}}_{\mathrm{i}\mathrm{j}}}{\underset{\mathrm{i}}{\max }{\mathrm{z}}_{\mathrm{i}\mathrm{j}}}} ,$$

(7)

Steps 2 and 3 together are zeroed unitisation with a veto threshold equal to 0 (cf. [104]) denoting gender balance.

Step 4. For each object, the determination of the value of the synthetic variable (taxonomic synthetic measure), which is a weighted arithmetic mean of the form [105]:

$${\mathrm{R}}_{\mathrm{i}}=\sum _{\mathrm{j}=1}^{15}{\mathrm{w}}_{\mathrm{j}}{\mathrm{z}}_{\mathrm{i}\mathrm{j}}^{\mathrm{*}} ,$$

(8)

where:

$$\sum _{\mathrm{j}=1}^{15}{\mathrm{w}}_{\mathrm{j}}=1,$$

(9)

• R—synthetic variable,
• R_i—value of the synthetic variable R for the i-th object,
• w_j—weight of the j-th normalised stimulant of feminisation of poverty.

The values of the weights depended on the number of diagnostic features related to each of the six SDGs included in the study; they are summarised in

Table 2. Weights for groups of normalised stimulants according to the 2030 Agenda SDGs.

	SDG 1	SDG 3	SDG 4	SDG 5	SDG 6	SDG 8
number of diagnostic features	5	1	4	1	1	3
weight of normalised stimulant	${\displaystyle \frac{1}{30}}$	${\displaystyle \frac{1}{6}}$	${\displaystyle \frac{1}{24}}$	${\displaystyle \frac{1}{6}}$	${\displaystyle \frac{1}{6}}$	${\displaystyle \frac{1}{18}}$

. By setting the weights in this way, each of the six SDGs was equally valid regardless of the number of diagnostic features representing it.

Step 5. Linear ordering of objects according to the decreasing values of the synthetic variable R. A higher place in the ranking (i.e., a lower rank value) indicates a higher level of feminisation of poverty in a given country compared to the other EU-27 member states.

Step 6. Classification of ranked objects into one of four classes of objects similar in terms of the level of feminisation of poverty:

$$\mathrm{c}\mathrm{l}\mathrm{a}\mathrm{s}\mathrm{s} \mathrm{I}:\overline{\mathrm{R}}+\mathrm{s}\left(\mathrm{R}\right){<\mathrm{R}}_{\mathrm{i}}\le 1$$

(10)

$$\mathrm{c}\mathrm{l}\mathrm{a}\mathrm{s}\mathrm{s} \mathrm{I}\mathrm{I}:\overline{\mathrm{R}}{<\mathrm{R}}_{\mathrm{i}}\le \overline{\mathrm{R}}+\mathrm{s}\left(\mathrm{R}\right)$$

(11)

$$\mathrm{c}\mathrm{l}\mathrm{a}\mathrm{s}\mathrm{s} \mathrm{I}\mathrm{I}\mathrm{I}:\overline{\mathrm{R}}-\mathrm{s}\left(\mathrm{R}\right){<\mathrm{R}}_{\mathrm{i}}\le \overline{\mathrm{R}}$$

(12)

$$\mathrm{c}\mathrm{l}\mathrm{a}\mathrm{s}\mathrm{s} \mathrm{I}\mathrm{V}:0\le {\mathrm{R}}_{\mathrm{i}}\le \overline{\mathrm{R}}-\mathrm{s}\left(\mathrm{R}\right)$$

(13)

where:

• $\overline{\mathrm{R}}$—arithmetic mean of synthetic variable R,
• s(R)—standard deviation of synthetic variable R.

EU-27 member states in class I have the highest level of feminisation of poverty, while those in class IV have the lowest.

3. Results

In the first stage of the empirical study carried out, the diagnostic features for men and women presented in Table 1 were transformed based on Formulas (3) and (4), determining the values of gender gaps (V_j). Selected descriptive characteristics of the transformed diagnostic features, i.e., gender gaps, are presented in

Table 3. Selected descriptive characteristics of gender gaps.

j	Mean	Median	Min	Max	Skewness	Number of Values of Gender Gap
						Vj < 0	Vj = 0	Vj > 0
1	1.8	1.5	−0.3	4.7	0.66	2	0	25
2	0.6	0.4	−0.6	2.6	1.12	4	1	22
3	0.4	0.4	−1.6	2.5	0.02	8	1	18
4	−1.6	−1.4	−7.7	1.2	−1.47	23	1	3
5	0.7	0.9	−0.8	3.1	0.59	6	1	20
6	0.7	0.5	−0.3	3.5	1.39	3	3	21
7	−3.2	−3.4	−8.6	1.9	−0.10	25	0	2
8	−7.7	−7.7	−17.8	3.5	0.07	26	0	1
9	−0.1	−0.2	−1.3	1.3	0.48	16	3	8
10	−2.6	−1.9	−13.6	0.4	−1.96	22	1	4
11	20.3	19.9	5.1	37.7	0.10	0	0	27
12	−0.1	0.0	−1.2	0.2	−2.35	11	13	3
13	5.9	4.3	0.3	19.2	1.15	0	0	27
14	9.7	8.3	1.7	19.7	0.64	0	0	27
15	0.2	−0.1	−1.3	5.3	3.54	15	0	12

.

The values of the V_j variables indicate the direction of the gender imbalance. From the information in Table 3, it can be seen that the three variables V₁₁, V₁₃, V₁₄ had only positive values, which means that in all EU-27 member states, in 2020, there was a disadvantage for women due to the diagnostic characteristics: inactive population due to caring responsibilities, young people neither in employment nor in education and training, employment rate—in every country, their values (in percentage) for women were higher than for men. The maximum gender gaps for these three features were the highest among all 15 variables, being respectively: 37.7 pp. (Poland), 19.2 pp. (Czechia), 19.7 pp. (Italy).

The values of the other gender gaps were both positive and negative and, in a few cases, equal to zero. The largest number (as many as 13) of values equal to zero indicating gender balance had variable V₁₂, meaning that in 13 countries, there were equal proportions of women and men having neither a bath, nor a shower, nor indoor flushing toilet in their household. Furthermore, in most of the remaining countries, men were at a disadvantage compared to women due to the previously mentioned feature and only in three countries (Austria, Czechia, Romania) was the situation reversed. The six variables V₄, V₇, V₈, V₉, V₁₀, and V₁₅ (i.e., all the characteristics included in the study concerning the quality of education and, in addition, the work, at-risk-of-poverty rate, and the long-term unemployment rate) had a preponderance of negative values (between 16 and 26) indicating that men were generally worse off in terms of these characteristics. It should be stressed that there were no variables representing gender gaps with all negative values, i.e., variables due to which men would always be disadvantaged. The closest to this was variable V₈—only in one country (Germany) was the percentage of female tertiary educational attainment lower than the corresponding percentage of males (by 3.5 pp.).

In the next stage of the empirical research, the transformed diagnostic characteristics (gender gaps V_j) were subjected to unitisation with a veto threshold denoting gender balance (Formulas (5)–(7)), during which stimulants of feminisation of poverty (Z_j) were determined. An aggregate synthetic variable (Formulas (8) and (9)) was then determined, on the basis of which the EU-27 member states were ranked and divided into four classes (Formulas (10)–(13)) according to the level of feminisation of poverty in 2020. The results obtained are presented in

Table 4. Values of the taxonomic synthetic measure, ranking and classification of the EU-27 member states.

i	EU-27 Member State	Ri	Rank	Class
1	Austria	0.2012	17	III
2	Belgium	0.1376	24	III
3	Bulgaria	0.2148	14	III
4	Croatia	0.1914	18	III
5	Cyprus	0.2426	9	II
6	Czechia	0.3655	4	I
7	Denmark	0.1004	25	IV
8	Estonia	0.3902	3	I
9	Finland	0.1443	23	III
10	France	0.2028	16	III
11	Germany	0.2680	8	II
12	Greece	0.5561	2	I
13	Hungary	0.2272	13	III
14	Ireland	0.2377	10	III
15	Italy	0.3014	6	II
16	Latvia	0.3193	5	II
17	Lithuania	0.1857	20	III
18	Luxembourg	0.1753	21	III
19	Malta	0.2367	11	III
20	Netherlands	0.0869	26	IV
21	Poland	0.2974	7	II
22	Portugal	0.2101	15	III
23	Romania	0.5598	1	I
24	Slovakia	0.2303	12	III
25	Slovenia	0.1739	22	III
26	Spain	0.1902	19	III
27	Sweden	0.0825	27	IV

.

The information in Table 4 shows that in class I of the EU-27 countries with the highest level of feminisation of poverty in 2020 in terms of the variables included in the study, there were four countries, namely Romania (1st place in the ranking), Greece (2nd), Estonia (3rd), and Czechia (4th). For these countries, most of the gender gaps and therefore also the stimulants of feminisation of poverty had positive values denoting women’s disadvantage, with six of them reaching their maximum. The highest level of feminisation of poverty in Romania was determined by the high values of most of the stimulants (higher than their average values)—as many as 10 of them had positive values indicating the disadvantage of women. In addition, Romania had the highest values of variables Z₇ (1.9 pp.) and Z₁₂ (0.2 pp.) among all the countries studied. Greece, which ranked second, had the highest values of the two poverty-inducing variables Z₃ (2.5 pp.) and Z₁₅ (5.3 pp.) among the EU-27 countries. Estonia and Czechia had the highest values of the variables Z₆ (3.5 pp.) and Z₁₃ (19.2 pp.). In Greece and Czechia, moreover, the largest number (as many as 11) of the stimulants of feminisation of poverty had positive values.

Five countries were in class II with a high middle level of feminisation of poverty: Latvia, Italy, Poland, Germany, Cyprus. For these countries, the six stimulants of feminisation of poverty reached their maximum. Latvia (5th) had the highest value of the Z₄ stimulant (1.2 pp.), Italy (6th)—Z₉ (1.3 pp.) and Z₁₄ (19.7 pp.), Poland (7th)—Z₁₁ (37.7 pp.), Germany (8th)—Z₈ (3.5 pp.), and Cyprus (9th)—Z₁₀ (0.4 pp.).

Class III, with a low middle level of feminisation of poverty, was the most numerous, with as many as fifteen countries in it. Only the last three stimulants of feminisation of poverty reached their maximum there: the highest values of stimulants Z₂ (2.6 pp.) and Z₅ (3.1 pp.) were for Bulgaria (14th place in the ranking) and Z₁ (4.7 pp.) for Lithuania (20th place). The last places in this class (21st–24th) were occupied by Luxembourg, Slovenia, Finland, and Belgium, respectively.

In class IV of the EU-27 countries having the lowest level of feminisation of poverty in 2020, three countries are ranked: Denmark (25th), Netherlands (26th), Sweden (27th). For Netherlands and Sweden, the taxonomic synthetic measures were similar, their lowest values being determined by the very small values of all stimulants, with as many as seven having values equal to 0.

The spatial distribution of the EU-27 country classes is shown in Figure 1.

The average value of the taxonomic synthetic measure was highest in class I with the highest level of feminisation of poverty at 0.468, in subsequent classes it decreased to reach a value of 0.090 in class IV with the lowest level of feminisation of poverty (Table 4). Similarly, in most cases, the average values of stimulants of feminisation of poverty were highest in class I and in the subsequent classes were increasingly smaller. The exceptions were variables Z₂, Z₈, Z₉, and Z₁₂, but the differences between their mean values in each class were insignificant. Some variables, especially those related to the quality of education, had the fewest positive values (denoting women’s disadvantage) of all the variables: variable Z₈ had only one value different from zero (for Germany), Z₇—two (Romania, Czechia), Z₄—three (Czechia, Latvia, Belgium), Z₁₂—three (Romania, Czechia, Austria), Z₁₀—four (Greece, Czechia, Germany, Cyprus). It can be seen that most of the countries mentioned belonged to class I and class II, only Austria and Belgium were from class III. The average values of the selected stimulants of feminisation of poverty in each class are shown in Figure 2.

4. Discussion

The relative poverty indicator AROP used in the empirical study (SDG indicator 1.20 from the Eurostat database) expresses the percentage of people at risk of income poverty with the cut-off point set at 60% of median equivalised income after social transfers. From its preliminary analysis, it can already be seen that in 2020 in the EU-27, women were worse off on average than men—17.4% of women were at risk of monetary poverty, while the corresponding percentage for men was 15.8%. However, the AROP indicator is not the only measure of poverty in Europe; in addition to it, Eurostat’s database on SDG 1 contains nine more indicators of poverty reduction (also linked to other SDGs) mainly from the Survey on Income and Living Conditions, EU-SILC [106]. Selected indicators from the Eurostat database on SDG 1 were used, among others, by Piwowar and Dzikuć in their study of poverty and social exclusion in the Visegrad Group Countries [107], Sompolska-Rzechuła and Kurdyś-Kujawska for the assessment of the development of poverty in EU countries [108], Bąk and Perzyńska in their study of poverty and social exclusion in EU-28 [109]. This paper proposes that indicators from the Eurostat database on different goals can be used to assess the level of feminisation of poverty in EU-27 member states.

According to the study, the highest level of feminisation of poverty in 2020 was in Romania, for which as many as 10 of the 15 stimulants of feminisation of poverty had values greater than their mean values, including the highest values among all countries surveyed for the differences between the percentages for women and men early leavers from education and training (1.9 pp.) and having neither a bath, nor a shower, nor indoor flushing toilet in their household (0.2 pp.). Although the latter value is not very large, it should be noted that only in two more countries besides Romania (Greece and Austria) was the proportion of women who did not have access to water and sanitation greater than the corresponding proportion of men—and this diagnostic characteristic is important as one of the causes of period poverty. The analysis of the SDG indicators (i.e., untransformed diagnostic characteristics) by sex further shows that Romania was also the country with the highest proportions of women with no access to water and sanitation (21.3%), early leavers from education and training (16.6%), and experiencing severe material and social deprivation (26.5%) across the European Union. The percentage of women at risk of income poverty after social transfers in Romania was the second highest among the EU-27 countries (24.6%). According to Eurostat data, the reduction in the percentage of the risk of poverty rate, due to social transfers (excluding pensions) in Romania was only 15.8% for women and this was the lowest value in the EU-27.

Other EU-27 member states having the highest level of feminisation of poverty in 2020 are Greece, Estonia, and Czechia. In Greece, representing the southern European model of social policy, unemployment benefits are low and people in need have to rely on family [98]. Meanwhile, the country had the highest gender gap among the EU-27 member states for women and men living in households with very low work intensity (2.5 pp.) and those in long-term unemployment (5.3 pp.). Analysis of the SDGs by sex further shows that Greece was also the country with the highest number of women of all EU-27 countries living in households with very low work intensity (13%), long-term unemployment (13.5%), and in addition, housing cost overburden (34.2%). At the same time, Greece had the lowest values for two variables that can exacerbate female poverty in the country: the employment rate for Greek women was only 48.7%, and only 71.5% of girls participated in early childhood education. The unfavourable high ranking of Greece and Czechia in terms of feminisation of poverty was determined by the high values of as many as 11 out of 15 poverty stimulants. Czechia additionally had the highest value in the EU-27 of the gap between the percentages for young women and men neither in employment nor in education and training (19.2 pp.). Czechia represents the Central and Eastern European model of social policy, in which social security benefits are the lowest [98]. Despite this, according to an analysis of SDGs by sex, only 11.7% of Czech women were at risk of income poverty after social transfers and this was the lowest value among all EU-27 member states. At the same time, the corresponding percentage for men was also the lowest (7.2%), making the gender gap in this respect one of the largest.

The Benelux and Scandinavian countries with very low values of the poverty feminisation stimulants were ranked last in the 2020 poverty feminisation ranking. Denmark, Netherlands, and Sweden had by far the lowest levels of poverty feminisation. These countries represent a citizen-supportive social-democratic model of social policy, with Sweden being the leading welfare state [109]. An important element of Sweden’s policy is the promotion of gender equality, enabling women to fulfil care and family responsibilities while striving for self-reliance and changing gender stereotypical choices [110]. The impact of such a policy is evident in the smallest percentage of women who are inactive due to caring responsibilities (7.5%) and the highest employment rate (77.4%) among the EU-27 countries. Analysis of the SDGs by sex further shows that Sweden and the Netherlands were the only countries among the EU-27 member states where neither women nor men reported having neither a bath, a shower, or an indoor flushing toilet in their household. In contrast, the percentages of women at risk of income poverty after social transfers were not the lowest in these countries, at 17.2% and 13.7% respectively.

Countries with the highest level of feminisation of poverty are mainly from the eastern wall of the EU-27. According to the classification of the International Monetary Fund, Romania is included in the emerging and developing economies, while Greece, Estonia, and Czechia belong to the advanced economies [111]. Countries in class II with a high middle level of feminisation of poverty are mostly advanced economies and even major advanced economies (Germany, Italy). The last places of class III and in class IV were exclusively advanced economies. This ranking of the feminisation of poverty on the basis of gender gaps largely reflects the ranking of the promotion of gender equality determined by the European Institute for Gender Equality on the basis of the Gender Equality Index values. According to the EIGE study on gender equality, in 2020, Sweden (which had the lowest level of feminisation of poverty) was the leader in promoting gender equality, while Greece (which was the second country with the highest level of feminisation of poverty) ranked last in promoting gender equality [69]. It should be noted that, since 2010, Sweden has had the highest and Greece the lowest progress in gender equality in the EU [112]. The distant places in the ranking of the promotion of gender equality in 2020 included Romania (25th), Czechia (22nd), and Estonia (17th), i.e., the countries with the highest level of feminisation of poverty. Germany was ranked only 11th, with a Gender Equality Index approximately equal to its average value for the EU-27 as a whole. The other countries with the lowest levels of feminisation of poverty Denmark (2nd) and Netherlands (5th) were high in the ranking of promoting gender equality in 2020 [69].

It is clear that a country’s level of feminisation of poverty is conditional on gender inequality (in this study expressed by gender differences in the values of diagnostic characteristics) and does not necessarily correlate with the level of poverty for women, let alone the level of poverty overall. As an example, Czechia and Sweden belonged to the classes with the highest and lowest feminised poverty levels, respectively, in 2020, while according to Bąk and Perzynska [109] and Sompolska-Rzechuła and Kurdyś-Kujawska [108] in 2018 and 2019 belonged to the classes with the lowest and medium poverty levels, respectively. In 2020, the relative poverty rate of AROP overall was 9.5% and 16.1% in these countries, and 11.7% and 17.2% for women, respectively. The highest AROP values were in Bulgaria (23.8% and 25.8%), which ranked only 14th in the feminisation of poverty ranking. The 8th place in the feminisation of poverty ranking of economically highly developed Germany, which was in the class with the lowest poverty levels in 2019 [108] and should, therefore, not be surprising. In 2020, the proportion of people at risk of income poverty after social transfers overall was 16.8% in Germany and the corresponding proportion of women was 16.1% (i.e., less than in Sweden). Meanwhile, in 2020, Germany was the only EU-27 country in which the proportion of women who had completed tertiary education was lower than the corresponding proportion of men (a difference of 3.5 pp.). Furthermore, in Germany, the percentage of women who were inactive due to caring responsibilities was more than 29 pp. higher than the corresponding percentage of men, and this was one of the highest values among the EU-27 member states. This was probably strongly influenced by the conservative social policy model and the patriarchal family model preferred in Germany, in which women do not work outside the home [109,113]. According to a report by the international humanitarian organisation Oxfam, the social welfare institutions operating in Germany try to reduce inequalities [114]. At the same time, benefits for parents or the unemployed do not always stimulate the poorest to increase their income from work [115].

Reducing poverty and inequality are the challenges of today’s world, meanwhile, in late 2019, the COVID-19 pandemic began to spread around the world, leading to increased poverty and worsening inequality. The greatest impact of the pandemic has been on the poorest people. Between 2019 and 2021, the average income of the 40% of the population who are poorest has decreased by 2.2%, while the average income of the 40% who are richest has decreased by only 0.5% [116]. According to the United Nations [117], the pandemic has stalled the unbroken trend of 25 years of poverty reduction, and the combination of its impact with the effects of climate change, the Russian invasion of Ukraine, rising inflation and slowing economic growth means that more and more people will live in extreme poverty [118]. Assuming trends continue, 575 million people will be trapped in extreme poverty in 2030 [119], including more than 340 million women and girls [120].

The COVID-19 pandemic slowed down positive trends towards gender equality. According to Oxfam’s 2022 report, women, alongside ethnic minorities and people in developing countries, are among the groups most affected by the increase in inequality as a result of the COVID-19 pandemic—when, with the rising cost of living, the poor get poorer, women get poorer the fastest [121]. In 2020, women were 1.4 times more likely to drop out of the labour market than men, and unpaid domestic work and care work took up three times as much of their time [122]. Compared to pre-pandemic projections, it will take 36 more years to close the gender pay gap, or as many as 136 years [123]. Unfortunately, due to missing data for Greece and Ireland, the gender pay gap value could not be used in this study. Available Eurostat data shows that in the other EU-27 countries in 2020, the percentage difference between the average gross hourly earnings of male and female paid employees ranged from 0.7% in Luxembourg to 22.3% in Latvia, i.e., in all these countries, women were paid less than men.

The European Union is regarded as a world leader in gender equality, but inequalities emerge to varying degrees in its member states [112]. Both women and men experience gender-based inequalities, but it is mostly women who face poverty because of it. The feminisation of poverty is a whole mechanism that makes girls and women more likely to fall into poverty than boys and men [124]. The worse economic situation of women is often a manifestation of the social roles imposed on them by the patriarchal family model but is also sometimes caused by violence or gender discrimination [125]. To quote Tahira Abdullah, “Poverty has a woman’s face” [126], and its feminine character is exacerbated by the constant pressure of norms stemming from culture, religion, or tradition. Women are affected by the gender pay gap, unequal intra-household distribution of resources [127], the burden of unpaid domestic work, single motherhood, caring for family members, and menstrual poverty.

Human rights violations are an important aspect of poverty—the marginalisation and social exclusion of poor people are widespread, leaving them feeling a lack of respect and dignity [128]. The repercussions of the feminisation of poverty are not only significant because of the disadvantage to women themselves, the feminisation of poverty has negative consequences that also affect their families. Women, being family managers of poverty, undertake various activities to supplement their meagre budgets [129], at the same time reproducing poverty by sharing it with the children they are raising [124]. The severity of women’s poverty should be taken into account and a better situation in the family and society should be ensured. The fight against poverty should be combined with tackling gender inequality in all aspects of life. Meanwhile, as noted by D’Adamo et al., academic interest in SDG 5 topics is almost the lowest of all the SDGs [5]. The United Nations emphasises the need to accelerate action leading to SDG 5 and thus warns that failure to prioritise gender equality could jeopardise the entire Agenda 2030 for Sustainable Development [120].

In 2020, the strength of European women was evident in education, especially in secondary and tertiary education, so they should be further supported in this, but at the same time, men should also be supported in this sphere. Women need support in finding employment if they want to work but are prevented from doing so because of the need to care for relatives. Equality policies should also include the elimination of the gender pay gap. Due to a lack of data, this variable was not included in the empirical study conducted, but the phenomenon of unequal pay for women existed in all EU member states for which data were available. It is therefore necessary to provide access to these data from the remaining countries. Data collection on menstrual poverty should also be developed. In addition, all public statistics characterising the population should be presented by gender, and the way they are measured should be standardised to ensure comparability of information from different countries. To assess and measure the feminisation of poverty, it is useful to use indicators that represent gender differences in many aspects of life, as proposed in this paper.

5. Conclusions

The feminisation of poverty, signifying the disproportionate over-representation of women and girls living in poverty, grew out of gender inequality and, on this basis, should be considered as a multidimensional concept related to different spheres of life. The multidimensionality of the phenomenon of feminisation of poverty resulting from gender inequality requires methods and data that take this aspect into account to assess it. This paper proposes not to directly examine the level of poverty itself; instead, the diagnostic characteristics of female and male poverty were transformed into gender gaps so that they show gender inequalities. When comparing the values of the designated gender gaps, it was noted that all variables included in the empirical study confirmed the existence of a gender imbalance in terms of poverty. Most of the diagnostic features indicated the feminisation of poverty, but in a few aspects, its masculinisation became apparent. In all EU-27 countries in 2020, a significantly higher proportion of women than men were inactive due to caring responsibilities, a lower proportion of women than men aged from 20 to 64 years were in employment, a higher proportion of girls than boys were neither in employment nor in education and training. In contrast, the worse situation for men was revealed in tertiary education—in all countries except Germany, the percentage of men in tertiary educational attainment was lower than the corresponding percentage of women. In most countries, with the exception of Romania and Czechia, the share of male early leavers from education and training was higher.

Undoubtedly, the feminisation of poverty that has grown out of gender inequality is a reality and not a myth. As the empirical study showed, this was the case, albeit to varying degrees, in all EU member states in 2020. High levels of feminisation of poverty were found in both countries whose economies are categorised as emerging market and developing economies (Romania) and as advanced economies (Greece, Estonia, and Czechia), while the class of countries with the lowest feminisation of poverty included only advanced economies (Denmark, Netherlands, Sweden). Regardless, all countries with the highest levels of feminisation of poverty had the lowest values of the Gender Equality Index indicating the greatest gender inequalities, while countries with the lowest levels of feminisation of poverty were leaders in promoting gender equality.

The article’s findings are intended to support the achievement of SDG 1 and SDG 5. Due to the cross-cutting nature of the phenomenon of feminisation of poverty, the study also used sustainability indicators from the Eurostat database for other SDGs. Limitations related to the availability of statistical data were encountered during data collection. For several countries, data on the gender pay gap were missing. This important variable was therefore not included in the study, even though an unfavourable gender pay gap for women existed in all EU member states for which data were available. Direct data on period poverty was also not used, as it is not widely collected and published.