Caritas’s Work for the Goals of Agenda 2030: A Study on the Services Provided in Campania

: The United Nations’ Agenda 2030 has established a series of Sustainable Development Goals to address global challenges, including poverty, food insecurity, access to education, and social inequality. In this context, charitable organizations such as Caritas play a crucial role in mitigating the negative effects of these challenges and promoting fair and sustainable development. This study aims to analyze prevalent needs among individuals seeking assistance from Caritas in Campania and examine how the organization contributes to achieving the Agenda 2030 Goals in the region. The statistical investigation techniques considered include tandem analysis a dimension-reduction technique, such as multiple factor analysis, and then a cluster analysis to identify similar groups of individuals. These exploratory data analysis methods have allowed for the identification of common needs, including food assistance, support for education, employment, and housing assistance. Subsequently, Caritas programs and initiatives aimed at meeting these needs and promoting sustainable development are explored. The results indicate that Caritas plays a significant role in addressing the urgent needs of the vulnerable population in Campania and contributes to the goals of Agenda 2030, particularly those related to poverty alleviation, immigration, health promotion, education, employment, and the reduction of social inequalities. This study provides an important perspective on the relevance and effectiveness of Caritas’s work in the context of Agenda 2030.


Introduction
In a world marked by constant change, characterized by increasingly complex socioeconomic challenges and growing disparities among populations, the need for transformative action has become ever more urgent.
In response to these pressing issues, the United Nations' Agenda 2030 emerges as a guiding light towards a more sustainable and equitable future for all.This ambitious action plan, adopted in 2015 by all 193 UN member states, sets forth 17 Sustainable Development Goals (SDGs) that encompass a wide range of critical issues whose main objective is the achievement of sustainable and inclusive development [1].
The definition of sustainable and inclusive development is highly debated in the literature.The classical definition of sustainable development was first given by the Bruntland Commission in 1987, which says that sustainable development is "the development that meets the needs of the present without compromising the ability of future generations to meet their own needs".
Inclusive development, on the other hand, refers to economic growth that ensures fair distribution across society and creates opportunities for all [2].For an in-depth reading on "sustainable development", see "The Sustainable Development Goals Report 2023 (United Nations)" [3].In this rapidly evolving global landscape, humanitarian and social assistance organizations like Caritas play a fundamental role in translating the ideals and objectives of Agenda 2030 into concrete and tangible actions at the community level.
Founded on the principles of solidarity, compassion, and love for one's neighbor, Caritas has been committed for decades to combating poverty and social injustice and providing support and assistance to vulnerable individuals, regardless of their social, religious, or geographical conditions.
The 17 SDGs are categorized into three main domains: the "social domain", focusing on poverty alleviation, reduction of social and economic inequalities, and improvement of access to education and healthcare; the "economic domain", aimed at fostering innovative economic growth, generating employment opportunities, and sustaining economic resilience; and the "environmental domain", dedicated to biodiversity conservation, mitigation of climate change, and preservation of ecosystems.
In this context, Caritas primarily contributes to the social domain by "promoting the testimony of charity, which is the concrete love for others.The dimension of charity permeates and enriches the life of communities" [4].
Through its global network, Caritas delivers a range of services from food support to healthcare, education, and vocational training aimed at improving the lives of vulnerable populations.Beyond providing assistance, Caritas mobilizes resources and fosters a culture of solidarity and active participation in society through awareness-raising projects, advocacy, and training.
In this paper, we focus on analyzing the characteristics of individuals who seek assistance from Caritas, examining how their needs and circumstances align with the broader goals of Agenda 2030.We explore how Caritas contributes concretely to realizing a more just, fair, and sustainable world for these individuals.To delve deeply into the impact and effectiveness of Caritas's initiatives in implementing Agenda 2030, with particular attention to the local context and social dynamics at play, we employ tandem analysis [5,6], which uses a dimension-reduction technique, such as multiple factor analysis, and then a cluster analysis to identify similar groups of countries.
Through this deep dive, we hope to offer new perspectives and valuable insights that can inform and guide future efforts towards achieving the SDGs and improving the well-being and dignity of all people, everywhere in the world.
This paper is organized as follows: Section 2 is a discussion of the theoretical framework; Section 3 presents the data; Section 4 introduces the data analysis methodology; Section 5 is an analysis of the results; and Section 6 provides the conclusion of this study.

Inequality and Poverty
In contemporary discourse, the intricate relationship between inequalities and poverty has garnered significant attention as researchers and policymakers seek to understand and measure these interconnected challenges [7].In the literature, inequalities have been studied from various perspectives.Smith [8] defined them as "differences in the distribution of income and wealth, as well as economic and social opportunities, among different groups within a society".
The linkages between inequalities and poverty are profound and multifaceted, shaping the lived experiences of individuals and communities.
However, the literature on poverty and economic and socioeconomic inequalities has greatly expanded over the years.The first distinction to be made concerns the difference between poverty and inequality.They are often discussed together, but they are distinct concepts with different implications.The former refers to the condition where individuals or groups lack the financial resources to meet basic living standards, such as food, shelter, and healthcare.It is an absolute measure that defines whether a person's income is below a certain threshold necessary for a minimum standard of living.It is important to note that the literature distinguishes between various levels of poverty, specifically between absolute poverty and relative poverty [9].The latter, on the other hand, refers to the uneven distribution of resources, opportunities, and wealth within a society.It is a relative measure that highlights the disparities between different segments of the population.Practically, poverty is about the absolute level of living-how many people cannot attain certain predetermined consumption needs.Inequality, indeed, is about the disparities in levels of living-for example, how much more is held by rich people than poor people [9].The UN recognizes the importance of combating poverty and reducing inequalities, including Goals in Agenda 2030, specifically in Goal 1 and Goal 10, called, respectively, "Eradicating extreme poverty for all people everywhere by 2030" and "Reduce inequality within and among countries" [1].
Caritas plays a crucial role in reducing poverty, aligning its efforts with the indicators outlined in Goal 1 of Agenda 2030 for Sustainable Development.One of the key indicators under Goal 1 is the proportion of the population living below the international poverty line.Caritas dresses this by providing direct financial assistance, food, and essential services to those in need, ensuring that vulnerable populations have access to basic necessities.
It should be noted that in the literature, there are various ways to measure the progress of Agenda 2030 [10].The literature distinguishes between measurement at the national, local, and urban levels [11,12].
However, the aim of this paper is not to analyze the extent to which Caritas contributes to poverty reduction but rather to study the characteristics of the individuals who seek assistance from Caritas.
To our knowledge, there is no work dealing with this topic.Agenda 2030 is an international plan, whereas our study focuses on the Italian context, specifically in Campania.We will now proceed with a description of the phenomenon of poverty in Italy, followed by an analysis of how Caritas contributes to addressing this issue.

Poverty in Italy in the Context of Agenda 2030
Focusing solely on Europe, it is clear that we are still far from achieving Goal 1 of Agenda 2030, which aims to reduce by 15 million the number of people at risk of poverty or social exclusion in Europe.
Eight years after the adoption of the Sustainable Development Goals and three and a half years into the socio-health crisis caused by COVID-19, we have experienced significant setbacks.Regrettably, the pandemic, the energy crisis, and the war in Ukraine are having a highly negative impact on our progress toward these targets.
Currently in Europe, more than 95 million people, comprising 21.8% of the population, are living in conditions of poverty and/or social exclusion.This percentage remains relatively stable compared to 2021 when it was at 22% [13].However, the impact of coronavirus is evident, as indicated by Figure 1, showing a reversal of the previously observed signs of improvement starting from 2020.Through a comparison among European countries [13], we can notice that in Italy, people at risk of poverty and/or social exclusion account for 24.4% of the population, which is higher than the EU average [13].
As it is not the objective of this paper to define the risk of poverty or social exclusion, it suffices to know that individuals in this category include those who live in at least one of the following situations: in households at risk of poverty, defined as having an income below sixty percent of the national median income; in conditions of severe material and social deprivation; in households with low work intensity.For further details, please refer to the literature (they are, for example, defined in the annual report on poverty by ISTAT [14]).If we shift from a European context to a national context, the southern part of Italy, known as the Mezzogiorno, emerges as the area most affected by poverty according to ISTAT data [14].

A Look at Data
Caritas is making a relevant contribution to the cause by conducting research and promoting concrete actions.For a comprehensive exploration of this topic, refer to the Caritas Report published in 2023 [15].In this context, we analyze a dataset concerning individuals who visited Caritas.The data are accessible on the Ospoweb platform, which is used by listening centers for entering data related to the beneficiaries of Caritas services.
The survey covered all the listening centers of the Diocese of Caserta.This latter, which encompasses almost the entire territory of Caserta and nine other municipalities, with a population of 210,000 inhabitants over an area of 182 square kilometers, includes 67 parishes.
The dataset consists of n = 1837 observations representing individuals who sought support from Caritas.In particular, the information was collected by Caritas staff from April to September 2023 and subsequently aggregated to form the dataset used for analysis.Various types of information were collected from these individuals, amounting to a total of p = 37 variables, as detailed in Table 1.In this table, the variables used in the analysis are highlighted in bold.Additionally, the first column contains the labels of the variables useful for interpreting the MCA graphs, while in the second column, detailed descriptions are provided.Further, the third and fourth columns pertain to categories.The categories column is the column where we describe the categories of the variables; the cells of this column are empty when they coincide with the categories label.For each variable, the nature of the variable and its respective categories (in the case of qualitative variables) have been specified.
A problem encountered during data analysis concerns the presence of missing values.There are several ways to handle missing data [16], among these, one can choose, when feasible, to perform imputation using specific techniques.However, when the data do not allow for this, for instance due to an extremely high rate of missing values (in our case, exceeding 2/3 for certain variables), one may opt to remove that variable.This is the approach we have chosen.
After making the changes, the dataset consists of 17 variables.For the needs of the statistical techniques used which we will describe later, transformations were performed on the quantitative variables.Specifically, such variables-i.e., DRES; NC; CHD; FC-were transformed into categorical variables, dividing them into non-empty and equidistant classes of values.
In the next paragraph, we will proceed with the description of the statistical techniques used.

Methodology: Tandem Clustering
For the multivariate analysis of our data, among the most widely used techniques proposed in the literature (such as multiple factor analysis for mixed data by Pagés [17] and nonlinear principal component analysis by Gifi [18]), we consider tandem clustering [5,6], which can be viewed as a method of minimizing redundancies in the data.
Here, tandem clustering uses a dimensional reduction technique, specifically multiple correspondence analysis [18][19][20][21], to create new variables that are uncorrelated and then applies cluster analysis to form classes using these new variables.
Starting from the factors extracted by multiple factor analysis (MCA), instead of using the original variables, the statistical units (the individuals who have sought assistance from Caritas, in our case) are grouped by using a hierarchical agglomerative clustering algorithm [22].As a result, only the most important variables will lead to the identification of similar groups of individuals.
Indeed, the factors, being orthonormal, have the advantage of providing the same impact on the (dis)similarity index used to measure the distance between the groups of units.The results of this integrated analysis approach allow us to better specify the needs of individuals and find confirmation of the individuals' profiles already identified in MCA.
It is common in the literature to use this type of technique for this kind of analysis [23][24][25].

Multiple Correspondence Analysis
The multiple correspondence analysis is utilized to explore the relationships among a group of categorical variables observed across a population of statistical individuals or units.By generating new variables (latent variables) and pinpointing an optimal low-dimensional space, MCA serves as a statistical technique for assigning scores to units and each category of variables.If categorical data have more than two categories, converting them into multiple binary variables is necessary.Each column of the starting matrix represents a variable that can take on values from one to the number of categories of the variable itself.However, on the table thus defined, the addition operations both for rows and columns would not make any sense.A first transformation is therefore obtained by defining a matrix in which each variable gives rise to as many dichotomous variables, i.e., taking on only the values 0 and 1, as there are categories of the variable itself.
Analyzing survey data by using MCA can be performed by calculating a super- variables observed on the same set of n individuals.Let D be the super-diagonal table of dimension J × J where the (k, k)th diagonal matrix contains the relative column marginal frequencies, p .jk = ∑ n i=1 x ij k n for the k-th variable.Observe that an indicator matrix implies coding the data in a complete disjunctive form [20,26].For example, the matrix X k consists of elements 0 and 1, where 1 represents that an individual/unit is classified into a category and 0 indicates that it does not share that characteristic.Therefore, the total number of categories under consideration is where the generic variable k has j k categories.There are many ways with which multiple correspondence analysis can be presented.One of those is to perform a (generalized) singular-value decomposition of the super-indicator matrix where Λ is the diagonal matrix of the singular values and U and V are the right and left singular vector matrices, respectively, which allow the computation of the coordinates for units and variable categories.The coordinates of the categories also allow one to display graphically the relationships existing among the variables.In particular, since each category is the center of gravity of the units (assisted by Caritas) that have chosen it, the proximity between two categories highlights those chosen by the same people or by very similar individuals: the proximity between two categories can therefore be interpreted in terms of association between them.Similarly, the proximity between two units allows one to highlight the (dis)similarity among people [27].

Hierarchical Agglomerative Clustering
To identify homogeneous groups of units, we perform a cluster analysis [28] on the unit coordinates obtained through MCA.In the literature [28], cluster analysis refers to a set of statistical techniques used to group statistical units based on the similarity of their profiles, described by a set of variables.The resulting groups of units should exhibit a high degree of internal homogeneity and a high degree of variability between the groups.Without prior knowledge of the appropriate number of clusters to analyze, we chose hierarchical agglomerative clustering from the many available classification methods.This method is summarized in the following steps.
Initially, each individual forms a separate cluster.In the second step, the two units with the minimum distance between them are merged.For calculating the distance, we used the Ward method, which is based on the decomposition of the total deviance into between-group deviance and within-group deviance.
Let c j be the jth cluster, n j its size, ēj its centroid, and ējj ′ the centroid of (c j ∪ c j ′ ).The Ward distance can be written as d(c j , c j ′ ) = n j d( ēj , ējj ′ ) + n j ′ d( ēj ′ , ējj ′ ). ( At each step, then, those two groups that obtain the minimum within-group deviance are merged.The third step calculates the distance between the new cluster (group) and all the other units.Finally, steps two and three are repeated until a configuration is reached where there is only one group.
The clustering process can be graphically represented through a dendrogram, from which it is also possible to read the aggregation index and appreciate how much a group is separated from the others.Of course, the aggregation index can be used in order to identify the suitable number of clusters: cutting the cluster tree after the fusion that corresponds to low values of the aggregation index and before those corresponding to high values of the aggregation index [29].Among the most common indices for assessing the goodness of clustering and determining the optimal number of clusters in the data, we apply the silhouette width [30].Let a i be the average dissimilarity between the observation i and all other points of the cluster to which i belongs.For all other clusters C, to which i does not belong, we calculate the average dissimilarity d(i, C) of i to all observations of C. Let b i = mind(i, C).The silhouette width of the observation i is defined by the following formula: Observations with an S i close to 1 are very well clustered.An S i around 0 means that the observation lies between two clusters.Observations with a negative S i are probably placed in the wrong cluster [22].

Results and Discussion
Let us proceed with the presentation of the results.The quality of representation of each category can be assessed by examining the contributions of the categories in Table 2.Note that if the sum of the first two dimensions exceeds 0.5, the label is bold.In general, the farther the categories are from the origin, the better the quality of the graphical representation.In Figure 2, the most characteristic categories are those farthest from the origin (which represents independence from the variables).The categories that most characterize the survey participants are indicated by warm colors, while the less significant categories are indicated by cool colors.On the right side of the image along the horizontal axis, there are variables and categories characterizing the first dimension.From this graph, we can start forming an idea about the associations of the modalities.On the right side of the figure along the horizontal axis, there are those individuals who come from a foreign country (Georgia, Iran, Santo Domingo, etc.) and with immigration issues; they are mostly young (see Table 2).The second dimension, the vertical axis, is characterized by those individuals who mainly reside in the province of Caserta, are homeless, and have turned to Caritas for economic issues (they received clothing and food parcels).
After conducting MCA, the research proceeds with the implementation of hierarchical clustering based on the coordinates of the individuals extracted by MCA.For selecting the optimal number of dimensions to consider, we chose to use the elbow method.Considering Figure 3, it is evident that from the second dimension onward, the eigenvalues stabilize, suggesting that we consider the number of dimensions preceding the flattening.
The implementation of hierarchical clustering based on the coordinates of the individuals extracted by MCA produces the dendrogram as shown in Figure 4.The dendrogram suggests that we consider three clusters.To confirm this choice, we used the silhouette index [22], which returned a value of 0.59, indicating that a structure in the data was found.Too see the characteristics of each cluster, refer to Tables 3-5.Below is a guide on how to interpret them.The first column of each table, titled CLA/MOD, represents the percentage of individuals in a specific category within a cluster relative to the total individuals who are in that specific category in the dataset.The second column, named MOD/CLA, expresses the percentage of individuals within a specific category relative to the total individuals in the cluster.The third column, indicated as Global, indicates the percentage of visitors within a specific category in the entire dataset.The fourth column displays the p-value, which represents the statistical significance of the category in each cluster.Finally, the last column provides the test value (v.test) related to the considered categories.

Clusters Description
In this section, we describe the three clusters, highlighting the characteristics that represent them.For each cluster, one or more primary needs for which they sought help were defined.
For each table describing the results, we highlighted in bold the most distinctive variables for characterizing the clusters.
The first cluster (Table 3) is mainly characterized by individuals with health problems (HP, v.test = 34.512)and economic issues (POV, v.test = 32.097).The services provided to them are material goods and services (v.test = 9.330).It is evident that these individuals are strongly characterized by being Italians (v.test = 11.617) and residents of Marcianise (v.test = 30.195),which could indicate a significant issue affecting this province.Regarding the second cluster (Table 4), it shows that the defining variables and categories are those related to work.Indeed, the individuals in this cluster have been listened to by Caritas and have sought support concerning poverty and employment issues (POV, v.test = 28.096);EI, v.test = 7.407).These individuals have reported living in houses with irregular contracts (unregistered rent, UR, v.test = 5.256).We can say that this cluster refers to employment and poverty issues faced by individuals residing in the province of Caserta, mainly in Maddaloni (v.test = 16.267),San Marco Evangelista (v.test = 7.062), and other areas.more assistance or to recognize problems that municipalities have not identified.In our specific case, the analysis highlighted three main issues: health problems, employment issues, and migration.Therefore, to address this problem and to reduce poverty and social inequalities, the province of Caserta should allocate more resources to these sectors or find alternative solutions.

Conclusions
The conclusions drawn from this study reflect a thorough analysis of the connection between the United Nations' Agenda 2030 and the organization Caritas, highlighting the crucial role the latter plays in implementing the SDGs and contributing to realizing a more just, fair, and sustainable world for all.Agenda 2030, with its 17 SDGs, provides a global framework for addressing complex socioeconomic challenges and promoting equitable and sustainable development worldwide.
Caritas, founded on the principles of solidarity, compassion, and love for one's neighbor, has been committed for decades for combating poverty and social injustice and providing support and assistance to vulnerable individuals, regardless of their social, religious, or geographical conditions.Through a wide range of programs and services, Caritas offers food assistance, shelter, medical care, education, and emotional support to those in need, thus contributing to the social domain of Agenda 2030.
One limitation of this study is the presence of categories with few observations.These categories were not grouped together initially because our focus was on capturing the unique characteristics of each individual.However, we are keen to further explore this issue by reapplying the analysis and grouping individuals with similar places of origin into broader categories.
Another avenue of development in this research involves assessing progress towards the 2030 Agenda goals.The literature has shown advancements in this area.Recognizing that individual municipalities also contribute to these goals, it would be beneficial to consider local indicators for monitoring progress in the province of Caserta rather than relying solely on a national metric [11,12].
The analysis conducted highlights the fundamental role of Caritas as a humanitarian and social assistance organization in translating the ideals and objectives of Agenda 2030 into concrete and tangible actions at the community level.Through its ongoing commitment and dedication to solidarity and compassion, Caritas remains a beacon of hope and a positive agent of change in the fight against poverty and social injustice, thereby contributing to a fairer and more sustainable future for all.From a statistical perspective, we considered tandem analysis due to its mathematical properties and the abundance of useful, interpretative graphical displays it offers.However, recent advancements in tandem analysis alternatives warrant investigation.These alternatives integrate dimensionality reduction and classification techniques simultaneously rather than sequentially.For handling qualitative variables, several promising techniques merit consideration, including MCA K-means [31], iterative factorial correspondence biplot [32], and cluster correspondence analysis [5].These methods, providing a powerful tool for simultaneous dimensionality reduction and classification, could enhance our analytical framework.

Figure 1 .
Figure 1.People at risk of poverty and/or social exclusion in the EU (incidence %)-Years 2015-2022.Source: Eurostat.

Figure 3 .
Figure 3. Eigenvalues in the elbow method.

Table 1 .
Description of variables.
PNMain need for which Caritas support is requested HI; DJ; ADD; FAM; HAN; EDI; IMM; EI; POV; HP; PRO Housing issues; detention and justice; addiction; family issues; handicap and disabilities; educational issues; migration/immigration issues; employment issues; poverty/economic issues; health problems; other problems RES Indicate the usual municipality of residence AFR; CA; CR; CV; CE; CC; CI; MA; MAR; MT; RE; SA; SF; SME; SNS; SMV; SMCV

Table 2 .
Contributions and Coordinates.