Labor and Inequalities: A Bibliometric Analysis of International Literature

—The objective of this research was to identify some characteristics of international scientific publications that address inequalities in the world of labor. The bibliometric study adopted the Knowledge Development Process-Constructivist (ProKnow-C) as an intervention instrument. In addition to the theoretical background that presents a synthesis of the main discussions and research results identified in the selected portfolio, the primary results include: a) the constitution of a bibliographic portfolio of 36 articles with scientific recognition, which presents different sub-themes related to inequalities at work; b) the indication of the most used keywords, which show the studies of gender, race, social class, and the phenomenon of discrimination at work as the most recurrent; and, finally, c) the indication of the main authors and journals prominent in these topics, both in the bibliographic portfolio and in their references.


I. INTRODUCTION
Inequalities in the world of labor and their possible causes are multifaceted and complex [1]. In the context of a globalized economy, it is possible to refer them to an international division of labor, roughly understood as the "specialization" of certain economic activities of countries, characterized as central and peripheral economies. Central economies, formed by the "developed" countries, focus on industrial activities that add greater scientific and technological potential and capacity for capital accumulation, as well as, consequently, better working and living conditions in their societies. Peripheral economies, composed of the "underdeveloped" countries, focus on economic activities involving the production and supply of food and raw materials, with lower added value and, as a result, worse working and living conditions [2].
Similarly, from a micro analysis perspective, inequalities in the labor market tend to be perceived (objectively) in a division of labor, in which a set of occupations and/or jobs (differentiated from each other) is produced, being distributed and rewarded, also, in a differentiated form between individuals and certain social groups [3]. Thus, the social organization of labor not only builds but reinforces and (re)produces the varied social inequalities [4].
The ideology of meritocracy has been commonly indicated as the social mechanism responsible for this differentiation. However, many studies with different approaches and epistemological perspectives have shown that there are other *Correspondence: michelfpolis@gmail.com variables responsible for inequalities in work that are not explained by individual merit or human capital [5,6]. These inequalities are, above all, explained by sociodemographic characteristics and social properties of individuals and social groups, such as social class, place of birth, ethnicity, gender, age, religion, sexual orientation, physical characteristics, etc. [7][8][9][10].
Thus, this study performed a bibliometry to know a little more about the "state of the art" of inequality in the world of labor, and the main research, sub-themes, approaches, authors, and prominent journals, etc., indicated by Demo [11] as the first path to be followed by the researcher after outlining the research theme, using the Knowledge Development Process-Constructivist (ProKnow-C) as an intervention instrument.
According to Ensslin et al. [12], the method proposes the selection of bibliographic portfolio according to the area of interest of the researcher, observing the established delimitations, with the composition of a portfolio of articles endowed with scientific recognition and aligned to certain research themes.
Finally, this article is structured as follows to achieve the proposed objectives: after this introduction, in the next section (Theoretical Background), without the pretense of covering the complexity of the phenomenon of inequality at work, we present a synthesis of the main discussions and some of the results of the international scientific publications identified in the bibliographic portfolio object of this research. The following sections present the methodological framework; an introduction to the process of database research and bibliometry; the procedures for selecting the theoretical reference and its results; and, finally, the final considerations.

II. THEORETICAL BACKGROUND
The systematic review of the literature on inequalities in the world of labor has shown that this is a plural and complex theme, typical of social phenomena. It is examined under the gaze of various disciplines, such as sociology [13], economics [6], geography [4], administration [14] etc.; approaches: qualitative [5], quantitative [3], and mixed [15]; research strategies, such as ethnography [16], case study [17], exploratory-descriptive studies [8], among others, and, consequently, different epistemological perspectives. The research on the subject also covers different levels of analysis, such as individual [18], group [7], organizational [19], etc.
These studies involve a set of sociodemographic aspects and properties of people and social groups that prove decisive for a "successful" insertion in the labor market, with an impact on the working and living conditions of the worker. In this sense, social class, place of birth, ethnicity, gender, age, religion, physical characteristics, etc., function as "social markers", distinctive marks, signs of distinction, or social stigmas [20].
Citing the study conducted by Weeden [21], Grand and Tahlin [3] highlight that inequalities in the labor market in Modern societies tend to be perceived (objectively) in terms of a) division of labor, in which a set of occupations and/or jobs (differentiated from each other) are produced; b) distributed; and c) rewarded, also, in a differently between individuals and certain social groups. Thus, the social organization of labor not only builds but reinforces and (re)produces the varied social inequalities [4].
In modern societies, the ideology of meritocracy, legitimized by the permanent acquisition of human capital (in the form of school capital, professional training, etc.), has been indicated as the one responsible for the differentiation in these aspects. However, many empirical studies have shown, with a wealth of details, that there are other variables (not explained by this route) that influence unequal relationships at work in addition to the qualification factor [5,6], [24], [18].
Snel et al. [9] understand that inequalities at work are perceived, predominantly, in terms of social class i and "complemented" by other "social cleavages", such as those of gender, ethnicity, age, sexual orientation, etc. The term "intersectionality" has been designated to try to explain the complexity of crossing multiple axes of differentiation (economic, political, cultural, psychic, subjective, and experiential) of individuals and social groups in specific social contexts ii [25].
Farris and De Jong [25] advocate the "lens of intersectionality theory" for a deeper understanding of the phenomena that generate inequalities in the workplace. For the authors, inequality is intersectional and can be analyzed in three dimensions (interrelated among each other): a) Structural, referred to a lasting and relatively shared social environment, which puts at a disadvantage less empowered individuals and groups, while favoring dominant individuals and groups; b) Institutional, fueled by policies and practices that produce discriminatory treatment in institutional environments (school and work environment, for example); c) Discursive, influenced by the structural and institutional dimensions, results in the intersection of discursive formations that (re)produce inferiorized images of people and social groups.
The literature is vast in indicating the various intersections that impact the work and life of some workers. For example, a study of young second-generation immigrant women, of North African and South Asian descent, in six member states of the European Union (Denmark, France, Italy, the Netherlands, Spain, and the United Kingdom), identified discrimination regarding dress code, religion, among others, which implies a disadvantage in the labor market, such as: difficulties of insertion, access to less qualified and stereotyped jobs regarding gender status, etc., disadvantages, in large part (beyond gender issues), arising from ethnicity and social class [25].
In line with these findings, studies on gender inequality in labor relations, in various sociocultural contexts, demonstrate the structural constraints of patriarchal relations and the penalties imposed on women, where the organization of production and labor tend to (re)produce gender inequalities (in addition to those concerning classes and castes) and deepen the vulnerability of working women, usually allocated in less qualified, invisible and low-paid jobs [26,27], [16].
Other groups still face discrimination in the labor market, such as the black population, indigenous peoples, people with obesity, people with disabilities, people of non-heterosexual orientation, etc., that, intersected with other "social markers", such as those of gender and social class, attenuate or aggravate the level of inequality they face.
Drydakis [8], found a significant difference in salary, with disadvantages for gay and bisexual men in his exploratory-descriptive study on income discrimination by sexual orientation (between gay/bisexual and heterosexual men), based on data from the Athenian labor market, isolating a series of variables, among them, qualification. The most "educated" gays and bisexuals face smaller pay gaps than gays and bisexuals with less schooling.
Foster and Wass [28], in their case study in the United Kingdom, explored the relatively neglected disadvantages of people with disabilities in the workplace, where positions designed around the notion of an ideal (non-disabled) worker have implied resistance by organizations to implement adjustments to include workers with disabilities.
Caliendo and Lee [7] studied the discrimination at work of people with obesity and overweight in Germany. According to the authors, obese women are the group that finds the most discrimination by the organizations. Despite applying for more job applications and engaging in more training programs, they have more difficulties entering the labor market. The group also showed significantly lower wages than people who were overweight or weighed within the "normality" standard.
Davis et al. [13] analyzed the phenomenon of poverty among Native Americans and its potential basis in labor market opportunities in metropolitan and non-metropolitan environments. Holmes [29] identified segregation in the (agricultural) workplace involving indigenous Mexicans. Poverty, low wages, limited work opportunities, poor housing and living conditions, with an impact on the health of this group, are the main problems identified in these works for this social group.
Brewster and Lynn [1] analyzed racial discrimination in consumer markets in the United States. The findings suggest that white and black clients discriminate against black servers by giving them lower tips. According to the authors, this unequal treatment is not related to the qualification of the server or the quality of his/her service. The possible causes remain undefined and are indicated for further research.
The status of migrant worker (whether national or international) has also received special attention in the literature on inequality at work. The decision to migrate may reflect a series of strategies (temporary or permanent): economic survival; refuge; acquisition of experiences, new professional skills; proficiency in the language of the host country, etc.
The inequalities experienced by international migrant workers are also multifaceted and depend on a number of variables that mitigate or worsen their working and living conditions in the country of reception: status, whether legal (with work authorization) or illegal (undocumented); levels of professional qualification and work experience; the sectors of the economy in which they are allocated; occupational levels; among others. Factors that, as it turned out, intersect with some "social markers", such as ethnicity (skin color), gender, dress code, accents, social class, etc.
Among the main problems faced by international migrant workers, the following stand out: Racialized perceptions related to visible "social markers", such as skin color and belonging, problems related to cultural differences, dress codes [30]; difficulty with the recognition of their qualifications and finding work compatible to that in the country of origin [5,24], [18]; barriers regarding language mastery [5,24]; experience underemployment, devaluation, and downward mobility, forcing entry into ethnic and gender markets [18]; modern servitude and slavery, as in the case of the exploitation of diplomatic domestic workers, aggravated by the restrictive bond and work authorization for a single family [31]; differences in working hours [14]; low wages [32]; institutional and practical discrimination, affecting opportunities for professional progress [4]; among others.
National migrant workers, likewise, suffer discrimination at work, mostly motivated by regional differences, especially by the rural vs. urban cleavage. Gagnon et al. [33], studied internal workers in China, in jobs with formal employment contract, and identified income discrimination between urban and rural migrants, with a greater extent of discrimination among rural migrants. Ma [6], identified "explained" and "unexplained" factors for income disparities when analyzing wage differences between urban and rural migrants and local residents. The "explained" factors include those relating to "human capital" (such as individual qualification levels). The "inexplicable" factors can be attributed to individual characteristics, such as place of birth.
The literature review has evidenced not only the problems faced by these individuals and social groups but the different strategies they use to circumvent the discrimination and exploitation they experience in their daily lives.
In the case of international migrant workers, studies indicate that some manage to improve their working conditions with longer residence and acquisition of professional experiences and proficiency in the language of the receiving country [15], achieving some economic mobility, especially in segments that already have experience (in the country of origin or in the country of reception). Above all, when the sector in which they are allocated is dynamic, they circumvent exploitation by acquiring new skills and seeking better working conditions with the help of their ethnic networks, as observed in the study conducted by Hagan et al. [32] on economic mobility of immigrant workers in the construction sector in North Carolina, USA.
Others manage to "circumvent" discrimination, as in hiring processes in the receiving country, omitting (in their resume) more qualified work experiences and "transliterating" personal names to the language of the country, as pointed out by Kosny et al. [5] in his study on Polish and Lithuanian workers in Australia.
Parutis [24] explored the work experience of Polish and Lithuanian immigrants in the UK, noting that these workers expect jobs that provide personal and professional development in addition to economic gains, and utilize different forms of cultural capital (such as new acquired skills and language proficiency of the receiving country) to increase economic capital.
Another strategy used to overcome the difficulties is the horizontal and vertical migrant networks, functioning as a kind of "cushion" for precarious conditions. Cultural and linguistic ties and the existence of ethical communities facilitate access to the labor market and mobility [34]. The return to the country of origin is also indicated as one of the coping strategies [32].
At the institutional level, some policies and practices have been used to address the phenomenon of inequality at work. In the governmental sphere, Kurtulus [22], in his descriptive exploratory study, underlines the fundamental role of government policies in the United States, the so-called "affirmative actions", to reduce the wage and occupational disparities (in positions of authority and high qualification) of women and black and Hispanic people.
Ortileb and Winterheller [14] demonstrated the relevance of institutional and cultural factors in promoting more egalitarian labor relations. Exploring the differences in the working hours between migrants and non-migrants in four European countries (France, Sweden, Austria, and the United Kingdom), the authors identified the preponderant role of the mode of regulation of the working hours and the "breadwinner models" iii of each country for equality or inequality at work. Huffman et al. [35], in his research on organizational policies and gender (wage) inequality, emphasize the relative importance of organizational policies to reduce income differences between men and women, highlighting their greater effect based on wage distribution, indicating that such policies do not benefit women uniformly, requiring specific policies for the highest strata of the wage pyramid.
Meyers and Vallas [19], in turn, emphasize types of "diversity regimes" (utilitarian or communal) in an exploratory study in worker cooperatives, reviewing the concept of Acker [36] on "inequality regimes" iv . The authors discuss how organizations can address and work with the issue of inequality differently (diversity regimes), obscuring the functioning of inequalities or promoting attention to their presence. Corroborating Acker's perspective, the authors will demonstrate, based on a case study, that the prospects for change are increased in environments where inequalities have become more visible (rather than invisible/naturalized) and contested.

III. METHODOLOGY
This section aims to present the methodological framework used to achieve the objectives of this scientific work to allow readers an understanding of its design and execution. Fig. 1 explains the methodological choices, presenting the assumptions that guided the path taken to achieve the objective of this study. The Process for Knowledge Development -Constructivist (ProKnow-C) was used as the intervention instrument for the operationalization of bibliometry.

IV. DATABASE AND BIBLIOMETRY RESEARCH PROCESS
According to Demo [11], the initial perplexity before a research theme is primarily overcome by the literature review. It allows the researcher to know what has already been studied or what has been studied on certain topics. It can, generally, be carried out through the survey of written publications and/or by electronic means [37]. With the evolution of information systems, the use of databases, that is, indexing systems of journals, books, theses, reports, event annals, among others, has become a more common means to facilitate the search for bibliographic references, especially when it is aimed at knowing what treatment a specific research topic has received in time. Therefore, as a starting point, researchers have resorted to bibliometric techniques to measure, interpret, and evaluate the results obtained in their searches [38].

V. PROCEDURES FOR THE SELECTION OF THE THEORETICAL FRAMEWORK
This section aims to explain the procedures used for the selection of articles, elaborate a theoretical framework, and build a bibliometric analysis of the portfolio of selected articles with the purpose of analyzing different aspects of scientific publications on inequalities in the world of labor.
According to Karlsson [39], the activities of literature analysis help the researcher: a) in obtaining scientific support since it is based on scientific production on the theme of their interest; b) in the justification of the choice of the theme and the contribution of their research project; c) in the foundation of the methodological framework; d) in outlining the scope of research; d) in the development of critical analysis of the literature and treatment of comprehensive and scattered information.
The selection process of the theoretical framework was divided into three stages, detailed in the next subsections, namely: a) preliminary investigation; b) selection of research portfolio articles; C) bibliometric analysis of the portfolio.

A. Preliminary Investigation
The procedures detailed below were carried out between May and July 2021. We used the Web of Science (or ISI) database to prepare the theoretical review on the topic of inequality in the world of labor, which resulted in a set of available articles to be considered in the selection process [40]. This database gives rise to the JCR (Journal Citation Report), journal impact factor. Thus, the authors of the research understand that this is one of the databases of greatest scientific contribution to the field of research in question today.
Once the database was defined, we chose the keywords that characterized the first filter for the article selection. With the proposed theme: inequalities in the world of labor, the researchers elected, a priori, the keyword "labor" as a term of one of the axes of the theoretical framework. For the purpose of combining the keyword "labor" to the theme inequality, another research axis was sought in the ISI database texts published from 2011 to 2021, using the terms "inequality", "discrimination", and "exploration" to compose a wider sampling. After an unstructured reading of titles and abstracts of the most cited articles related to the keywords, we observed the adequacy of these terms, focusing on the research in development. Although they are different terms, "discrimination" and "exploration" maintain a close relationship with the phenomenon of inequality experienced by people and social groups at work. Fig. 2 indicates the combination of keywords used to compose the initial mass of references for the selection process.

B. Selection of Research Portfolio Articles
The search activities were carried out between May 19th International Journal of Social Science and Humanity, Vol. 13, No. 2, April 2023 and 30th, 2021. In addition to the keywords (Fig. 2) and the date of publication (from 2011 to 2021), we also used filters, such as: "Document type": articles; and "Web of Science categories": Business, Management, Social Science, Economics, and Multidisciplinary, the latter for believing in the relevance of an interdisciplinary approach to a better understanding of the phenomenon in question. The database search resulted in 5,529 references, as illustrated in Fig. 3.   We used the EndNote application to manage the references [41]. This tool allowed us to identify 95 references to be excluded from the sample due to duplicity. The library consisted of 5,434 references after the exclusions. We read the titles of these 5,434 articles to observe their alignment with the present research. After this analysis, 3,800 references were excluded because they were not aligned with the research. Thus, 1,634 references remained to be analyzed (Fig. 4).
The 1,634 references that aligned with the research theme through reading the titles were analyzed for their "scientific recognition" since their publication. This analysis consisted of all references consulted through the Google Scholar (2010) regarding the number of citations and then ordered in descending order.   With this information, the authors of the research in question established a cut-off value for the most cited articles.
International Journal of Social Science and Humanity, Vol. 13, No. 2, April 2023 This cut-off value was defined from the generalization carried out by Juran [42] to the postulate of Pareto [43], in which a small minority of the population accounts for most of the effect. Contextualizing this theorem to this research, it is understood that, if the research selects a percentage of the most cited articles, these represent the majority of the scientific recognition present in the current set of articles [44]. Thus, the cut-off value presented a selection of the most cited references, with a percentage of up to 80.53% of all citations obtained by the 1,634 references analyzed so far. Summing all citations of the 1,634 references analyzed, we obtained 34,314 citations. Thus, the articles individually cited 24 times or more represent 27,634 citations, that is, 80.53% of all citations of the 1,634 references selected so far. Thus, the cutoff point adopted was 24 citations or more concerning scientific recognition. With this identification of the cut-off value, 342 references were selected by the number of citations, as presented in Fig. 5. It is important to emphasize that the 1,292 least cited references will still undergo an analysis process under other criteria, with the possibility of being part of the final portfolio.
Once the references with the highest scientific recognition were selected, the analysis was carried out regarding the alignment of the abstract to the focus of the research in question. Of the 342 abstracts analyzed, 302 were excluded for not presenting the phenomenon of inequality (in its different manifestations) in the world of labor as a central discussion, according to the research design. Thus, as explained in Figure 6, there were 40 articles that: a) align with the reading of title and abstract; b) have a relevant volume of citations; c) have an accessible abstract.
These 40 articles with scientific recognition and aligned with the research theme were selected to compose the theoretical framework on inequalities in the world of labor. However, further analysis is required at this stage of the research. Two conditions were defined when evaluating the 1,292 least cited references to enable some inclusion of these references in the final portfolio: a) articles published less than two years before the analysis since they did not have the possibility of being well cited yet; b) when published in more than two years, these must be authored by some researcher already present in the group of 40 articles aligned with the abstract and with scientific relevance.
With the definition of these two assertions, 316 articles of the 1,292 articles analyzed were published between 2020 and 2021. Of the 976 articles that were published before 2020, only 13 articles are by authors present in the portfolio of the articles already selected. Thus, of the 329 articles selected for the reanalysis process, five were selected after reading their abstracts, given their alignment with the research object. Fig.  7 illustrates the reanalysis process and explains the number of articles that have passed in each activity of the article selection process. With these reanalysis procedures of the least cited articles, five were selected and added to the 40 previously selected articles, totaling 45 articles for the final portfolio.
As a final procedure, the 45 articles were then read in full, in order to assess their adherence to the research topic. Thus, of the 45 previously selected articles, 9 works were excluded due to misalignment with the research topic.

C. Bibliometric Analysis of the Portfolio
The bibliometric analysis of the portfolio was divided into three stages: a) bibliometric analysis of the selected articles; b) bibliometric analysis of the references of the selected articles; and c) classification of articles and journals according to academic relevance in the sample.
Three aspects were highlighted in the bibliometric analysis of the selected articles: a) scientific recognition for the number of citations (Chart 1); b) number of articles per journal (Fig. 9); c) the main keywords used in the portfolio (Fig. 10); and d) number of articles per author. For this last aspect, the researcher Palenga-Möllenbeck was the only author with two articles selected for the final portfolio [45], [46]. The other authors had only one article selected.  Regarding the number of citations, the studies conducted by Parutis [24], Dancygier and Laitin [47], and Tomaskovic-Devey [48] stand out. The first two works bring the issue of immigrant labor and discrimination in the labor market as a central discussion [24], [47], and the study conducted by Tomaskovic-Devey [48] reflects on the phenomenon of inequality in the workplace.
The most prominent journals in the portfolio are Sociology, Social Problems, International Migration, and Ethnic and Racial Studies, with two articles each. The analysis of the keywords highlights the main themes studied in the bibliographic portfolio when the subject is inequality at work, highlighting the studies of gender, race, social class, and the phenomenon of discrimination. However, when considering term variations, interest in the phenomenon of (i)migration of individuals and social groups is prominent.
When addressing the bibliometric analysis of the references of the 36 selected articles, 1,271 references were cataloged. Of these 1,271 references, we analyzed: a) prominent articles by the number of citations in Google Scholar with adherence to the theme (Fig. 11); b) journals of greater prominence (Fig. 12); and c) authors of greater prominence, according to citation in the references (Fig. 13).   Among the studies highlighted in the references of the bibliographic portfolio, according to the number of citations in the Google Scholar, are: Ronald Oaxaca [49], on income differences between men and women in the urban labor market; Joan Acker [50], on organizational structure and gender differences; and Alan S. Blinder [51], reflecting on the different types of discrimination based on income differences between white and black men and between men and women. The journals highlighted in the publication in the references of the BP are the American Sociological Review, Work and Occupations, and Work, Employment, and Society. Of the authors most cited in the references stand out: George J. Borjas, Bridget Anderson, and Li Shi. The themes of (i)migration and discrimination in the labor market are predominant among the studies conducted by these authors.
We defined two analysis axes regarding the classification of articles according to the relevance in the sample: a) number of citations in Google Scholar that the article has obtained since its publication; and b) number of citations in the bibliographic references of the articles in the portfolio. Fig.14 graphically presents the two dimensions defined to classify the articles according to their academic relevance and those highlighted in the BP. The quadrants were defined to leave only five articles in evidence, approximately 15% of the sample. Two of these articles are located only on the first axis [24], [47], two on the second axis [48], [46], and one between the two axes [6].

VI. CONCLUSION
The objective of this article was to identify the characteristics of international scientific publications on inequality at work, through the application of the systematic review method ProKnow-C [52].
For this purpose, we searched the Web of Science (or ISI) database to select a bibliographic portfolio (BP), aligned with the theme, relevant, and recognized by the scientific community, contemplating the bibliometric analysis of 36 articles (and their references).
As results, in addition to the synthetic reflection on the treatment that the theme has received in the literature, sub-themes, main discussions, and some of the results and findings of the research, under the perspective of different disciplines, identified in the BP, we can highlight: a) the indication of the most used keywords that show the most recurrent themes when the subject is inequality at work; b) the appointment of the main authors (and their works) who address the different themes, both of the BP, and its references; and c) the identification of the most prominent journals in these themes, of the BP, and of its references.
Concerning the keywords, studies of gender, race, social class, and discrimination (at work) stand out. However, studies aimed at understanding the phenomenon of (i)migration of individuals, and social groups are prominent when considering variations in terms related to keywords.
Among the main authors and works with scientific recognition in the BP, are Violeta Parutis [24], Rafaela Dancygier and David Laitin [47], Donald Tomaskovic-Devey [48], Helma Lutz and Ewa Palenga-Möllenbeck [46], and Xinxin MA [6]. These studies highlight the intersection of topics, such as (i)migration, inequalities, discrimination, and gender. Among the works that stand out for the number of citations in Google Scholar in the references are those of Ronald Oaxaca [53], Joan Acker [50], and Alan S. Blinder [54]. Studies of gender, race, and discrimination predominate. The authors most cited in the BP references are, respectively, the American economist George Jesus Borjas, the sociologist Bridget Anderson, the economist Li Shi, and the also economist Xinxin Ma.
Among the most prominent journals in the publication of themes related to inequalities in work in the BP are Sociology, Social Problems, International Migration, and Ethnic and Racial Studies. The American Sociological Review, Work and Occupations and Work, Employment and Society stand out in the references.
Despite the results listed here, we suggest future research continue this study with the expansion of other databases that aggregate publications from other realities and contexts, especially Latin American. We also emphasize that the conclusions of this study are referred to the perspective of the researchers since each researcher defines his/her research criteria, giving him/her the cut, he/she considers pertinent to his/her interests.