Research Landscape on Hidden Workers in Aging Populations: Bibliometric Review

Abstract

In this study, we employ ‘hidden workers’ as a key concept to integrate the three vulnerable subgroups of aging workers: underemployed, unemployed, and discouraged workers. (1) Background: The challenges faced by underemployed, unemployed, and discouraged workers in the older population are complex. It would be beneficial to visualize the intellectual landscape of these three distinct groups in aging populations to understand which aspects have been highlighted by various disciplines and where gaps exist. (2) Method: Through a scientometric analysis of more than 50 years of research, this study identified the size, scope, and structure of knowledge on hidden workers in an aging population using 2831 articles collected from the Web of Science database in January 2024. (3) Results: Indeed, the multidisciplinary nature of hidden workers goes beyond welfare and labor economics and involves issues such as health, occupational science, behavior change, policy interventions, and circles around the keyword of unemployment. Keyword co-occurrence and co-citation analysis confirm that the spectrum of research on hidden workers is being carried out distinctly within distinct disciplines across the broader aging research horizon. (4) Conclusions: The relatively scattered and uneven intellectual, conceptual, and social landscape of research on hidden workers in aging populations evidently falls short of providing concerted policy recommendations for the population group. This study provides a conceptual understanding of hidden workers in different research clusters and identifies gaps and opportunities for future research.

1. Introduction

The growing trend of aging worldwide has resulted in an aging workforce (ILO 2022). As people live longer, they are more inclined to continue working. This can be viewed as a natural response to extended life expectancy and the need to maintain the financial stability of social security systems. The average retirement age in many countries is increasing, and there is a need to support the health and work ability of older employees to ensure an adequate workforce. In this context, researchers and practitioners have been promoting “Healthy and active aging”, “active participation for a lifetime”, “lifelong active societies”, and “multistage work lives” to emphasize the importance of productivity in an aging workforce (WHO 2021; Harasty and Ostermeier 2020; Scott 2021).

However, not everyone is entitled to a healthy and active later life. There are hidden workers who wish to be but are not in the workforce. Fuller et al. (2021) defined hidden workers as a grouping of three broad occupational status categories: missing hours (working one or more part-time jobs but willing and able to work full-time), missing from work (unemployed for a long time but seeking employment), or missing from the workforce (not working and not seeking employment but willing and able to work under the right circumstances) (Fuller et al. 2021). Their access to the labor market has long been recognized as a critical social problem, which may lead to public health concerns such as mental health issues and limited healthcare access. Hidden workers have been researched in various disciplines and portrayed distinctively in diverse disciplines, as shown in Figure 1.

Since the beginning of the welfare state (Esping-Andersen 1990), the older population has been at the center of attention regarding the magnitude of social welfare expenditures and labor market barriers. The aging population has been associated with various policy measures, namely, income safety nets, retirement age, employer-provided benefits, and subsidies, which have been implemented and examined in different countries (Hudson 2012).

A group of researchers have created a distinctive sub-discipline of ‘retirement’, in which retirement is viewed as a social structure, transitioning process, and dynamic experience (Szinovacz 2003). They study individual/group-level transitions in and out of the labor market and navigate old-age insurance and other regulations that impact their work, livelihood, health, and lifestyle without clear demarcation. At the intersection of retirement age and changing patterns of later working lives, various disadvantaged population groups have been highlighted within the older population as facing distinct challenges related to seeking work (Moen et al. 2022; Becker and Fiske 2022).

Public health as a discipline recognizes the importance of addressing the relationship between health conditions and the working life of the older population. Studies have shown that work-related factors can significantly impact the health and well-being of older adults in the workforce (Noguchi et al. 2022). Moreover, the quality of job experiences plays a vital role in extended workforce participation among older adults, demonstrated by job quality factors such as employment preferences and workloads influencing optimal health outcomes among older workers (Welsh et al. 2016). On the other hand, structural ageism, which includes discriminatory policies and practices in societal institutions, can contribute to workplace discrimination against older individuals, which is detrimental to their health and well-being (Chang et al. 2020). In other words, many of those who have the potential to work but are not currently working because of barriers, such as ageism, economic or social inequalities, illness, or acquired or congenital disability, cannot equitably access the health benefits of work (ILO 2022).

Scholars using the life course approach discuss how accumulated patterns of inequality and inequitable access to opportunities affect older job seekers, particularly noting that those with disadvantaged childhood experiences may potentially lead to greater isolation and limited access to job opportunities (Sakkeus et al. 2023). Freitag and Hill explored the tendency among older workers, particularly those belonging to racial minorities, to exit the workforce before age 65. These workers are less likely to have completed their education, have lower incomes, and are more likely to depend on disability-related income sources (Freitag and Hill 2021).

Research in work and employment studies indicate that older unemployed individuals may experience discouragement and stop searching for full-time jobs (Nivorozhkin and Nivorozhkin 2020). This is concerning because the job-finding probability of older unemployed workers is lower than that of prime-age workers (Flek et al. 2020). Moreover, an employer’s attitude and workplace support toward an ‘age-friendly’ environment continue to be critical for the recruitment and retention of the older population (Nagarajan and Sixsmith 2023).

It is evident that while various disciplines portray complex challenges associated with hidden workers, it would be beneficial to understand those challenges in a comprehensive manner. Hidden workers may be exposed to most of the risks and challenges identified by different disciplines and quite possibly experience those struggles altogether at the same time. Addressing these challenges requires an interdisciplinary approach (Nagarajan and Sixsmith 2023; Lee and Kang 2024). Despite the widespread acceptance of work-related challenges in aging populations, interdisciplinary knowledge exchange has been limited (Lee and Kang 2024). Therefore, this study aims to map the intellectual, conceptual, and social landscape of research on hidden workers in aging populations to contribute to an interdisciplinary understanding of hidden workers.

One potential reason for the limited impact on these populations is the lack of coherent terminology, which may limit the accumulation and impact of the academic literature. Thus, this study uses the term ‘hidden workers’ to combine the wealth of works on this topic. By using the term ‘hidden workers’, we can integrate the sporadic development of the literature on hidden workers across various disciplines.

To the best of our knowledge, based on our research and analysis of the related literature, there has been no systematic or bibliometric review of hidden workers in the aging population. Integrated knowledge of hidden workers in the aging population is expected to be useful in influencing policies and interventions aimed at improving later working lives. Such an understanding will not only help guide the urgent policy roadmap on hidden workers but also maximize research impact via aligned research and policy tractions.

The ultimate goal of this study is to provide a holistic and systematic overview of the research on hidden workers during the last decade to inform past and emerging structures in the research landscape. In doing so, this study will evaluate global progress and quantitatively assess the current research trends in hidden workers in an aging population. This comprehensive scientometric analysis and visual overview of the intellectual landscape of hidden workers addresses the following:

(1)

Summarize significant publication patterns among hidden workers in aging populations in terms of year of publication, subject category, journals, countries/regions, and institutes.

(2)

Evaluate conceptual and intellectual mapping of the hidden workers in different research clusters by using keyword co-occurrence analysis.

(3)

What are the implications for future research?

To provide answers to the questions about the broader subject of hidden workers, a scientometric analysis was carried out using the Clarivate Analytics’ Web of Science (WoS) database and science mapping tools, as explained in the following section. The goal of this study is to offer insights that contribute to the current understanding of hidden workers in the aging population through descriptive, analytical, and reflective questions.

2. Materials and Methods

Science mapping is a technique that involves visualizing research fronts, knowledge structures, or intellectual bases to gain insight. In order for these visualizations to be meaningful, particularly for the research questions mentioned above, a thorough research design is necessary to analyze the intellectual, conceptual, and social structures present in the literature.

Bibliometric Analysis-Science Mapping Approach

The study uses science-mapping research, with the goal of fostering theoretical and empirical growth in literature, by conducting a scientometric analysis on the topic of hidden workers in the aging population. By employing this approach, a comprehensive overlook of works from researchers from diverse disciplines is mapped to provide the organic framework and intellectual knowledge base of unseen workers within the aging population.

Science mapping uses the idea of bibliographic coupling, which has been widely used since Kessler introduced the concept in 1963 to describe the segmentation of research and to analyze the development of emerging areas in the scientific landscape (Kessler 1963). Small (1973) then introduced the co-citation method, which heralded the use of analytical citation tools, and others have developed various techniques for citation analysis (Liu 1993). In addition, co-authorship and collaboration networks have been employed in bibliometric analysis to explore the social structure of scientific works using a multidimensional scaling approach (Kruskal 1964), which has been applied to report structural features of social space.

Search terms were carefully put together by the research team and the La Trobe University librarian. The terms used in this analysis are listed in

Table 1. Search terms.

Concept	Search Terms
Older	old OR older OR aged OR elder* OR vulnerable OR “over 45” OR “over 50” OR “mature”
Hidden Worker	NEAR/3 (jobseeker* OR unemploy* OR “part time” OR underemploy* OR “under employ*” OR “seek* work*” OR “bridge employ*” OR “discourage*” OR “hidden” OR “invisible”)

* denotes the inclusion of all the interactions of words that share the same prefix.

below.

To this end, this study uses quantitative science mapping techniques and reports on a bibliometric analysis of hidden workers in an aging population published from 1966 to 2022, as well as a content analysis of the 25 most-cited articles. Specific consideration is given to the volume of articles published, the use of keywords over time, citation practices, categorization and visualization of disciplinary venues, and the topical focus of research clusters.

The Clarivate Analytics Web of Science (WoS) database was used to collect bibliometric data for analysis in this review, which were then translated into the BibTeX file. Web of Science offers a subject categorization scheme called ‘research areas that cover five broad disciplinary categories (Clarivate Analytics 2020): Arts and Humanities; Life sciences and Biomedicine; Physical sciences, Social Sciences; Technology.’

“All Databases” in Web of Science was used due to its interdisciplinary global representativeness of core scientific output and citation-analysis capabilities. To build the reference database, we identified all references related to all three hidden worker categories in the aging population using the keywords, ‘aging’, older’, or ‘old’. The electronic search was limited to original articles, review articles, and conference abstracts that contained at least one of the two identified groups of keywords in the titles, abstracts, or keywords.

Bibliometric analysis was conducted using descriptive, relational, and qualitative indicators. Descriptive measures account for the volume of publications and citations. Relational indicators refer to the extent of collaboration between authors and citation practice. Qualitative indicators were based on the keywords provided for each selected article.

A general content analysis was performed based on keyword searches to identify prominent conceptual domains characterizing the corpus of publications with cross-cutting themes among the older population. However, quality appraisal or deeper analysis should be conducted in the systematic review of meta-narrative reviews, as indicated in the research protocol registered in the OSF at osf.io/2adqk.

3. Results

3.1. Evolution of Research Landscape on Hidden Workers among Older Population

3.1.1. Historical Development

In January 2024, there were 2831 papers on hidden workers in the aging population in the Science Citation Index Expanded, Social Sciences Citation Index, and Arts and Humanities Citation Index databases, distributed across different document types. There were 2326 research articles which comprised 92.72% of the total production, followed by proceeding papers (186, 6.54%), reviews (61, 2.15%), early-access articles (43, 1.52%), and book chapters (26, 0.92%). Publications of all other types, such as editorial materials, meeting abstracts, letters, book reviews, and news items, were removed from the analysis.

The first such article was published in 1966. However, it took 30 years before modest productivity would flourish. The results showed that research on hidden workers has experienced steady growth over the last two decades. Approximately 50% of all papers have been published in the last 10 years. Figure 1 represents this trend in an emerging and growing research area of hidden workers in the aging population.

The trends in annual publications on the research topic from 1966 to 2023 indicate that the development process of hidden workers in aging populations can be divided into three stages according to the growth pattern of the literature. The first stage was the infancy period from 1966 to 2000, the second stage was the slow development period from 2001 to 2010, and the third stage was the rapid growth period from 2011 to 2023. During the past few decades, the number of WoS papers on hidden workers in aging populations produced in the 1980s was less than 100 and increased to almost 112 in 2012. By 2012, the number of WoS articles on hidden workers was almost 1000. Few researchers paid attention to this topic before 1990, based on the number of publications. After 2001, the annual number of WoS publications grew exponentially, indicating that research on this topic has recently gained more attention. This information is provided to reflect on the disciplinary and geographical distribution of studies to be reported in later sections.

3.1.2. Disciplinary Evolution of Publication

Based on the categorization of subject areas in the Journal Citation Report’s WoS, the scientific production data of hidden workers in the aging population throughout the past century were spread across 168 subject areas.

The topic area was mainly located in the fields of public, environmental, and occupational health, psychiatry, and economics, as shown in

Table 2. Periodic publications classified by WoS subject category.

Web of Science Categories	1966–2000	2001–2010	2011–2023	Record Count	Ratio
Public Environmental Occupational Health	16.36%	24.90%	30.4%	453	15.94%
Economics	26.06%	24.66%	16.8%	417	14.67%
Psychiatry	15.15%	13.17%	15.73%	249	8.76%
Medicine General Internal	6.06%	8.38%	9.87%	147	5.17%
Sociology	9.70%	5.27%	3.47%	98	3.45%
Industrial Relations Labor	9.09%	5.03%	2.67%	88	3.10%
Education Educational Research	3.03	5.43%	3.47%	86	3.03%
Management	7.06%	5.43%	2.93%	85	3.00%
Environmental Sciences	0	3.51%	10.4%	83	2.92
Gerontology	9.09%	4.22%	4.27%	83	2.92

. The table indicates that research in the key disciplines began relatively early in the economics discipline, and increasingly more studies have been conducted in the public health and occupational health disciplines. Psychiatry has remained an important discipline that focuses on the mental health of older unemployed individuals or job seekers.

When recategorized in terms of Web of Science categories, the disciplinary clusters become more apparent. Public Environmental Occupational Health ranked first with 453 records (15.939%), followed by Economics as the second-highest category with 417 publications (14.673%). If occupational health science, psychiatry, and medicine are grouped into greater health disciplines, the proportion increases to almost 30%. While the dominance of the economic and health sectors is prevalent, the health sector has been growing more rapidly in terms of disciplinary concentration. Economics as a discipline is still strong but shows a clear tendency to decline. Environmental science might be linked to studies closely related to occupational health science, where proximal and decimal environments are believed to be relevant and impactful to all human endeavors and processes in occupation.

3.1.3. Core Journal of Publication

Table 3. Top 25 journals classified by total publication number and proportion.

Source	Record Count	Proportion
Plos One	50	1.76%
BMC Public Health	43	1.31%
International Journal of Epidemiology	40	1.41%
Applied Economics	22	0.77%
BMJ Open	22	0.77%
Social Science Medicine	17	0.60%
Social Psychiatry and Psychiatric Epidemiology	16	0.56%
Journal of Affective Disorders	15	0.53%
European Journal of Public Health	14	0.49%
Sustainability	14	0.49%
Social Policy Administration	12	0.42%
Frontiers in Public Health	11	0.39%
Labor Economics	11	0.39%
Journal of Epidemiology and Community Health	10	0.35%
Journal of Social Policy	10	0.35%
BMC Health Services Research	9	0.32%
International Journal of Equity in Health	9	0.32%
Journal of European Social Policy	9	0.32%
Regional Studies	9	0.32%
Social Policy and Society	9	0.32%
Scandinavian Journal of Public Health	9	0.32%
Work Employment and Society	9	0.28%
Aging Society	8	0.28%
BMC Psychiatry	8	0.28%
Cureus Journal of Medical Science	8	0.28%

shows the top 25 journal articles among the 1.709 journals in diverse disciplines. By observing the record count and proportions of each source’s publications, it becomes apparent that disintegration and divergence are the key features of this structure, which mainly results from the lack of central venues in journals. This is supported by the fact that the most productive journal, Plos One, had 50 publications (1.759% of all publications), potentially owing to its interdisciplinary nature. BMC Public Health recorded the second-largest number of publications, reaching 43 (0.513%) of the total articles.

This means that the scientific papers published on this research topic may potentially exhibit a lack of cohesion among the researchers involved in their publication. This also indicates that the intellectual structure of the study may display a scattered distribution. The high ranking of interdisciplinary journals in this field could be due to the unclear disciplinary position of this topic of research. While the older population hidden from the labor market has been studied from various disciplinary angles, the different layers of arguments and evidence might undermine the accumulated wealth of research to achieve the desired social impact.

3.1.4. Spatial Distribution of Publications

At the country level, Figure 2 shows spatial visualizations of the volume of publications worldwide. Scholars have emphasized research areas in 69 countries. The most productive countries are the USA, with 634 documents (17,433 citations), followed by England (259 documents), and Germany (212 documents). The next group of countries with over 100 publications were Australia (171 documents), the People’s Republic of China (131 documents), Canada (115 documents), Spain (111 documents), and Sweden (108 documents). A co-authorship analysis was performed for authors from 126 countries. Among the countries, 69 were compiled with the threshold of having at least five documents.

Figure 3 shows the network of co-authors on the topic across countries, which involves seven clusters. The sizes of the nodes determine the strength of the links. A country with a larger node represents more significant interactions and collaboration with researchers in other countries. The USA node size was the strongest node, with a total link strength of 31,996. England had half the total number of documents compared to the USA, but the total link strength was close to that of the USA (24,091). Germany and Australia ranked third and fourth, with 212 and 171 documents, respectively. The number of Canadian and Swedish publications was smaller than that of the aforementioned countries, but the total link strengths were 12,396 and 12,038 respectively, showing a relatively strong cooperative research strength.

The first network cluster was led by Australia and New Zealand together with Asian countries, including China, and the second cluster was led by Germany, formed by other European nations. The third cluster comprised South American countries, including Spain and Brazil. The third cluster was the UK cluster, which included Greece and Uganda. The fifth cluster was led by the USA, Canada, and the Netherlands, including Ethiopia and Iran. The sixth cluster comprised Northern European countries, including Finland and Sweden. These findings indicate that England, Canada, and Sweden are active in creating a collaborative network with researchers from other countries.

The overlay visualization from the VOSviewer software (Figure 4) illustrates the periodic concentration of research clusters by country. Publications from the USA, England, and Germany are displayed with purple bubbles, indicating relatively old publications from the region. Spain, South Korea, and Poland are the second-oldest groups in green bubbles. Finally, India, China, Kenya, and Nigeria are represented by yellow bubbles, indicating more recent publications. In short, network analysis reveals that developing countries are emerging networks in this area of research.

The analysis also found that a total of 3583 institutions published on the topic of hidden workers in aging populations. The top 20 most productive institutions on this topic and the total number of publications per institution are provided in the Appendix A. Among the 20 institutions, six came from the USA, four came from Australia, and three from England and Sweden. London University was the largest contributor of publications on this topic, accounting for 2.85% (81 documents) of all publications. Harvard University was the second most prolific institution, with 39 publications (1.372% of the total number of publications). The third most prolific institution was the University of Melbourne, with 38 publications.

3.2. Conceptual and Intellectual Development of Research Clusters on Hidden Workers

3.2.1. Keywords Occurrence Analyses

To determine which categories among the hidden worker categories were used and how their use evolved over time, we plotted the number of publications identified by each keyword against the year of publication. Figure 5 shows that unemployment appears to be the most frequent term used, and under-employment is also in use. However, discouraged workers have not been widely studied in the literature. Finally, there was an increased interest in ‘depression’ as well as in topics related to ‘mental health’, ‘COVID-19’, and ‘quality of life’, judging by the relatively higher ranking of these keywords. They represented the research emphasis on hidden workers in an aging population (Figure 5).

3.2.2. Keywords Cluster Analyses

We also examined the co-occurrence relationships among the top 90 high-frequency author keywords and co-keyword networks visualized using VOSviewer. The top three most frequently used keywords were again ‘unemployment’, ‘employment’, and ‘depression’, as shown in Figure 6. The map of author keywords in Figure 4 shows that ‘older workers’, ‘risk factors’, ‘COVID-19’, ‘disability’, ‘retirement’, ‘gender’, and ‘suicide’ also appear as frequent keywords. The countries also appeared with research associated with their research interests. Germany appeared with the keywords, human capital, social security, and retirement. In contrast, South Africa was associated with suicide, alcohol consumption, and HIV infection.

Author keywords appearing in articles that refer to hidden workers in aging populations were calculated, and the top 90 author keywords were used and clustered using VOSviewer from 1966 to 2023. Of the 5910 keywords, 1185 appeared twice, and 616 appeared three times. Only 295 keywords appeared five times. For keyword mapping, 90 keywords appeared ten times. The high percentage of only one author’s keywords probably indicates a lack of continuity in research and a wide disparity in the research focus. Another reason is that these keywords may not be standard or widely used in different disciplines. For instance, ‘job searching’ and ‘coping strategy’ may be standard keywords in work and occupational psychology literature, but not in other disciplines, despite the fact that they may be discussing similar content.

3.2.3. Co-Citation Analysis

To explore co-citation clusters across different disciplines, as shown in Figure 7, authors with a minimum of 20 citations were selected. Of the 61,318 authors, 201 met this threshold. As seen in the figure above, there are two compact and well-articulated central cores in the citation network: one centered around economic sector-led works and the other around health sector-led works. Top 25 publications on hidden workers are listed in Appendix B. Below is the description of the four key clusters identified from the co-citation analysis.

Group 1 (Orange) is a central cluster connected to other clusters and is interdisciplinary in nature. It includes works by Jahoda (1979, 1981, 1982) and McKee-Ryan et al. (2005), which share disciplinary roots in psychology. Nonetheless, the cluster expands to work and employment studies, where the decision-making process of individuals and socio-economic and health consequences fall upon them. Jahoda had progressed the field with life course perspectives, associating the phenomenon of unemployment with earlier life determinants of unemployment. Warr and Jackson (1984), from an occupational psychology perspective, examined long-term unemployment and its association with health measures and financial strain. Wanberg (2012) joined the stream of psychologists who had begun to contribute to the study of the factors associated with re-employment speed and quality. His work suggested that unemployment leads to a decline in psychological and physical health and an increased incidence of suicide. Hammarström and Janlert (2002) investigated the long-term effects of early-life unemployment and concluded that it may constitute a significant public health problem. While the work of Hammarström and Janlert (2002) is coded in red, the two clusters are closely related, where the publications in red are more likely to focus on public health rather than psychology.

Group 2 (Red) examines unemployment explicitly using a public health lens. Roelfs et al. (2011) conducted a systematic review of publications focusing on the association between unemployment and all-cause mortality among working-age persons. Bartley’s (1994) work provides a review of four mechanisms from unemployment to health: (1) the role of relative poverty, (2) social isolation, (3) loss of self-esteem, and (4) health-related behavior and the spectrum of scarring effects of unemployment. Wanberg (2012) continued the stream of work by collecting evidence on the individual experience of unemployment, its impact on psychological and physical health, and its association with an increased incidence of suicide. Martikainen and Valkonen (1996) conducted a prospective study to follow up on the mortality ratios among the unemployed in Finland. Those who remained unemployed had higher mortality rates than those who were re-employed. Stuckler et al. (2009) evaluated active labor market programs across EU states by analyzing the association between unemployment and suicide rates. They concluded that all-cause mortality rates were positively associated with unemployment; every US$10 per person increase in investment in active labor market programs reduced the effect of unemployment on suicide rates.

Group 3 (Green) comprises a broad spectrum of research in terms of disciplines. This includes the works of Fergusson et al. (2001) and Murphy and Athanasou (1999), which are more closely related to public health aspects as well as unemployment research from the perspective of social policy and welfare states (Esping-Andersen 1990, 1999). The welfare state stream of research follows up on the challenges of retirement decisions (Anxo et al. 2012; Tatsiramos 2010), unemployment insurance (Lammers et al. 2013; van Berkel 2020), incapacity benefit (Lindsay and Houston 2011), and social security (Engels et al. 2017; Reichlin 2019), which were also discussed in this cluster. Country-level longitudinal studies are popular methods for analyzing the type of welfare state or evaluating specific policies and targeted interventions. The works of Beatty and colleagues (Beatty et al. 2000; Beatty and Fothergill 2005) are coded in light blue but are closely related to arguments about welfare states, pointing out how the measurement of unemployment is distorted by the way that ‘sickness’ is defined and counted by social security systems. People in less advantageous areas, or more specifically those in closed-down mines, tend to apply for sickness benefits instead of unemployment benefits, which often have mutual obligations to keep applying for jobs. When faced with grim job prospects, individuals settle for sickness benefits, only to find that they would no longer be entitled to them due to the UK’s more stringent policies toward those recipients.

Group 4 (Blue) has a disciplinary foundation in economic analysis and is situated between public health and welfare state clusters. Chan and Stevens (2001) demonstrated that job loss has large and lasting effects on future employment probabilities, based on their examination of workers aged 50 and above who have experienced involuntary job loss. Kanfer et al. (2001) studied the relationships between various sociodemographic variables and individual characteristics in job search behavior and employment outcomes, providing policy implications for disadvantaged job seekers. Sullivan and von Wachter (2009) used administrative data on the quarterly employment and earnings of Pennsylvania workers in the 1970s and the 1980s to estimate the impact of job displacement on mortality. Their findings showed that, for male workers with high seniority, mortality rates in the year following displacement are higher than expected under normal circumstances.

In the introduction, disciplinary concepts related to hidden workers were introduced. Based on co-citation cluster analysis, it is worthwhile to depict the distance and overlap of different research clusters in identifying hidden workers. The simplified overview of Figure 8 illustrates the body of scientific research clusters related to hidden workers in aging populations and indicates the dominance of the welfare economy and public health clusters. The life course approach shares its disciplinary roots with the public health sector and impacts other research clusters. The relatively new stream of research closely linked to organizational and occupational psychology clusters is within work and employment science. The discipline of occupational psychology has been deeply invested in understanding the relationship between mental health and unemployment. This stream of research has been closely linked to work and employment studies to understand individual/interpersonal industry responses to (un)employment and job-seeking behaviors. Another emerging stream of research is interdisciplinary research on retirement. This cluster has its roots in both the welfare economy and public health sectors and aims to understand the dynamic nature of retirement and its individual/societal impact. Despite being an overly simplistic representation of the findings, Figure 8 illustrates the diverse portrayal of hidden workers and the urgent need to establish a more comprehensive and interdisciplinary understanding of the group.

4. Discussion: Uneven Research Landscape

One noticeable trait in the evolution of the research landscape is that the literature is dominated by the keyword, ‘unemployment’, as seen in Figure 5. Underemployment has been highlighted relatively recently, where discouraged workers or jobseekers among the older population have not made it into the top 90 high-frequency keywords. There are only a few studies that discuss discouraged workers, or those giving up opportunities in the labor market, from the occupational psychology discipline that evaluate individuals’ choices of retirement or expectations in the job search process.

This means that while we have a wealth of knowledge on how unemployment impacts individuals and society, there is still limited understanding of other groups that make up the hidden workers group. We are yet to understand comprehensively who the underemployed or discouraged workers are, why they remain hidden from the labor market, and what this means for an aging society. Not to mention that in current national statistics, discouraged workers are typically categorized as inactive labor forces, as they do not actively seek employment and are often identified as pension recipients or retired workers.

The next sign of imbalance is the dominance of the Economics and Public Health disciplines. Due to the ongoing international focus around the keyword ‘unemployment’, there have been internationally driven networks of research reflecting the problem of an aging population in relation to the labor market and public health. Nonetheless, these dominant disciplinary roots gave rise to newly emerging interdisciplinary clusters, such as the retirement cluster and the work and employment cluster. The retirement cluster, in particular, explores how segments of the older population face retirement pathways with respect to their attitudes and expectations toward the labor market.

The final but critical trait is the geographical concentration of work in Western societies. A total of 9 out of the 20 most productive institutions are American institutes. In conclusion, it is evident that universities and academic institutions in the United States are leading the production, collaboration, and communication of research in this area. Nonetheless, more recent works are arising from developing countries, especially the ‘emerging economies’. Nigeria, Egypt, Keyna, and Thailand are displaying promising developments in terms of the diverse representations of hidden workers.

5. Future Research Potential

There are several potential areas for future research regarding hidden workers among the older population. One such area is the development of interventions specifically designed for underemployed and discouraged older workers. This is a particularly complex task, as it involves factors such as country-specific barriers related to later-life employment and pension rules. More contextual information is needed when addressing the challenges faced by long-term unemployed older workers seeking to enter the workforce for the first time or return to work after an injury or illness. Designing an intervention for these groups would require an interdisciplinary understanding of their needs and circumstances.

Another area for future research is the study of dynamic transitions between the active and passive/inactive labor force among the aging population. While the transition from employed to unemployed (retired) or retired to employed is relatively well-studied among older populations, there is still much to be learned about the different subsets of hidden workers, such as discouraged workers and transitioning retirees. This research could provide important policy implications to further understand why some transitioning retirees become discouraged by the labor market and how they move within and beyond the hidden group.

6. Limitation

It is crucial to emphasize that our bibliometric analysis did not enable us to evaluate the quality of the publications. Moreover, our study encountered constraints in identifying the traits of various research groups within the hidden workers’ research in the aging population. For instance, in different research clusters, health may be understood and examined as a determinant, process, or outcome variable for hidden workers. To establish a potential collaborative space for researchers, a more rigorous understanding of the underlying paradigms is needed. Thus, future efforts are needed to describe the paradigmatic features of different research clusters on hidden workers.

7. Conclusions

In this research, we conducted a bibliometric analysis to assess the current status of hidden workers in aging populations. Our findings indicate a significant increase in the number of publications on this subject over the past 50 years, particularly in the last 20 years. A total of 2831 research articles were published in 1709 journals across 169 WoS subject categories. The primary areas of focus for this research include public health, environmental and occupational health, labor economics, community health, and aging studies. However, the publications are dispersed across numerous journals, with the highest concentration (2.5%) in interdisciplinary journals such as Plos One and BMJ Open. As a result, there is a lack of core journals dedicated to these topics, which may impede knowledge dissemination and exchange in this field.

Publication output is distributed unevenly across countries, institutions, and authors, with OECD nations accounting for the majority of articles. Additionally, the USA, England, Germany, Australia, China, and Canada play a significant role in international collaborative networks, fostering the creation, exchange, and sharing of knowledge on the subject. Sweden also holds importance in the Northern European and American collaborative network category, showcasing great progress in paper research production, citation, and cooperation, as well as overall strength in publications and research networks. The research landscape related to hidden workers highlights the central theme of unemployment among older populations, bridging various disciplines. Although the public health implications of unemployment are a key discipline in the topic, analyses of keyword co-occurrence and co-citation indicate potential opportunities for collaborative research in different research clusters. This study offers a comprehensive understanding of hidden workers across different research areas and identifies gaps and opportunities for interdisciplinary research on the subject.