Bibliometric Overview of Organizational Legitimacy Research

In recent years, the evaluator’s perspective of legitimacy theory appears to be a hot topic. We provide an overview by conducting a bibliometric review in this field. Based on the data during 1995 to 2020 from the Web of Science Core Collection database, this study investigates the present status, the evolution of the evaluator’s perspective, thematic change, and intellectual turning points via HistCite, CiteSpace, etc. Results demonstrate that the evaluator’s perspective is experiencing exponential growth. Suchman MC is the most cited author; While Palazzo G is the most productive. Academy of Management Review ranks first in citation, and Journal of Business Ethics is the most productive. The University of Lausanne enjoys the highest citation, and Copenhagen Business School publishes the most articles. The USA ranks first in terms of citation and publication. Author collaboration network is not close, while institutional collaboration and national collaboration networks are relatively dense. Highly-cited articles show the evolution of the evaluator’s perspective, and five evolution paths are found. The main areas are concentrated on signaling theory, business ethics, CSR communication, reputation, environmental disclosure, etc; 14 references act as intellectual turning points. The findings provide a panorama for researchers interested in this topic and help to better understand future research direction.


Introduction
Interest in organizational legitimacy has been growing considerably in recent years, especially in the fields of economics, business, and management (Bucheli & Salvaj, 2018;Suchman, 1995). Legitimacy is of great significance to firms and governments since they are viewed as principal bodies in production and political activities in the whole social system. Specifically, the legitimacy of a firm in the eyes of a government or society determines its survival and the resources available. If a firm is legitimate, resources or favorable policies from government and society can be expected (Oliver & Holzinger, 2008), while a lack of legitimacy may harm a firm's performance (Stevens et al., 2015) and even threaten the firm's license to operate and existence in a country (Henisz et al., 2014). Moreover, the legitimacy of a government in the eyes of social actors determines not only the stability of the government itself but also the order and stability of the society; more importantly, it may indirectly determine the legitimacy of a firm that has close ties with the government (Bucheli & Salvaj, 2013). Therefore the ability to acquire and maintain legitimacy is an essential condition for firms and governments to survive and succeed in a country (Darendeli & Hill, 2016;Swaner, 2017).
Legitimacy is defined as a generalized perception or assumption that the actions of an organization are desirable, proper, or appropriate within the socially constructed system of norms, values, beliefs, and definitions (Suchman, 1995). Firm legitimacy has been explored from different perspectives, such as institutional perspective, strategic perspective, etc. Institutionalists regard legitimacy as the internal need of the organization and focus on the importance of the organizational field in shaping institutional isomorphism (Deephouse & Carter, 2005), or institutionalization of new organizational structure and practice (Johnson et al., 2006). Legitimacy is viewed at the collective level, highlighting environmental determinism and the passive response of the organization (John & Lawton, 2018). Strategic scholars regard legitimacy as an important resource (Zimmerman & Zeitz, 2002) and investigate legitimacy at the organization level, stressing the active response of the organization and the role of managerial agency (Tost, 2011). They mainly address, for example, how organizations acquire, maintain, and repair legitimacy (Suchman, 1995;Zimmerman & Zeitz, 2002), how organizations lose legitimacy in case of crisis and environmental jolts (Sine & David, 2003), or how organizations respond to institutional pressure (Xie et al., 2020). Another stream of scholars has explored governmental legitimacy itself. Considering that the government is viewed as the maker of policy, institutions, rules, and regulations as well as the representative and defender of a given social order, the legitimacy of a government may dramatically influence the legitimacy of firms embedded into it. Government legitimacy is defined as the perceived recognition and support of political belief and value, political institution, economic and social policy, political conduction behavior, and its outcomes in political science (Almond et al., 2008). The maintenance, consolidation, obsolescence, and loss of government legitimacy are explored (Almond et al., 2008;Dietrich & Winters, 2015;Christensen & Laegreid, 2020).
However, the above-mentioned perspectives mainly concentrate on the receiving end of legitimacy, while ignoring the rendering end of legitimacy-evaluators. Even though legitimacy is ultimately a collective-level phenomenon, it's the outcome of individual-level legitimacy. Recent studies divert the scholars to individual-level legitimacy by introducing the evaluator's perspective. The individual's judgments and perceptions act as the "micro-motor" (Powell & Colyvas, 2008), which will coalesce to constitute collective-level legitimacy (Bitektine, 2011). As is widely recognized, legitimacy ultimately exists in the eyes of the beholder (audience, evaluating audience, or evaluator) (Zimmerman & Zeitz, 2002). Legitimacy can be viewed as actors' perceptions of the organization, as a judgment about the organization, or as the behavioral consequences of the two, demonstrated in actor's actions-"acceptance," "endorsement," and so on (Bitektine, 2011). Organizations of any kind will receive judgment and evaluation from different evaluators.
Literature of this perspective deals with, for example, who the evaluating audience is, which features of the organization are evaluated (Suchman, 1995), or how the judgment is formed (Hayes & Newell, 2009). The legitimacy of an organization in the eyes of evaluators can mainly be categorized into two categories, namely government-perceived legitimacy (e.g., legitimacy with regulators) and other social-actor-perceived legitimacy (e.g., legitimacy with investors, or advocacy groups, or employees, etc.). Evaluators interact with an organization by perceiving different aspects of the organization (e.g., size, quality, etc.). Finally, the judgment is concluded through the form of cognitive legitimacy or sociopolitical legitimacy (Bitektine, 2011). The evaluator's perspective thereby provides insight into social and cognitive factors, which is of great significance to improve an organization's legitimacy. Despite the growing interest in the evaluator's perspective, analysis of the development status of the topic is limited.
Several literature reviews on organizational legitimacy arise in recent years. A pivotal review by Deephouse and Suchman (2008) deals with the definition, dimension, source, and strategies of legitimacy. Based on this review, another review by Deephouse et al. (2017) expands the basic concept of legitimacy. Besides, Suddaby et al. (2017) discuss the concept, source, and strategy from treasure, process, and perception perspective. In addition to the above qualitative reviews, the quantitative review by Díez-Martín et al. (2021) identifies the intellectual structure of organizational legitimacy. While they focus more on a comprehensive review of legitimacy and organizational types are not distinguished in the above study. Scholars may not systematically assess the amount or the quality of previous research from the evaluator's perspective. This justifies the need for our study.
Our study aims to fill the gap by conducting a bibliometric review on organizational legitimacy from the perspective of evaluators. In addition, only firms and governments are selected to be the objects of legitimacy. We regard firms as the key object of legitimacy. Considering government legitimacy may influence the legitimacy of firms that have close relations with it, we include governments as a secondary object of legitimacy. These distinguish our study from the previous literature review.
We tend to figure out: (1) research status; (2) evolution of evaluator's perspective; (3) thematic structures; (4) intellectual turning points. Bibliometric analysis is a useful statistical method to meet the above research objective. This study conducts a bibliometric analysis based on the data during 1995 to 2020 from the Web of Science (WoS) database. Our analysis provides evidence for the significant role of the evaluator's perspective on organizational legitimacy and enriches the existing literature. This study may shed light on researchers who are new and interested in the evaluator's perspective of organization legitimacy so that they can get an overview and easily know which point is more helpful for future study. We hope that our analysis might serve as a basis for the study on organization legitimacy. This paper is organized as follows. Section 2 introduces the research method and data. Section 3 presents the results, followed by the discussion and conclusions section.

Method
Bibliometrics provides an approach to analyzing literature from an objective and quantitative perspective and is frequently used to organize information in a certain thematic field. To figure out the influence of certain characteristics and main areas in this perspective, citation analysis and cocitation analysis are adopted.
Citation analysis is a common method in bibliometric analysis. It provides a way to describe the development of a specific subject, and evaluate the influence of several characteristics including publications, authors, institutions, and countries. It's a useful means to construct a general picture in a certain scientific area (Merigó et al., 2016).
Co-citation analysis is a useful instrument to figure out the main areas of research in a certain field (Díez-Martín et al., 2021). Co-citation can measure the similarity of two articles by the times they're cited together. Articles being cited together may form different clusters, from which we can grasp the thematic structure of a certain topic.

Data
We construct the data collection according to the data source, search strategy, and data processing techniques and tools.
Data source. The data are retrieved and downloaded from the WoS Core Collection database with "Social Sciences Citation Index" editions. Serving as the gold standard for research discovery and analysis (Su et al., 2019), WoS provides access to multiple databases and complete information for bibliographic analysis (Guo et al., 2019).
Search strategy. The term "government* legitima* or soci* legitima*" and the field tag "Topic" (including title, abstract, author keywords, and Keywords Plus) from 1995-01-01 to 2020-12-31 are selected. The year 1995 is chosen as the starting year because the first record of the topic appeared in 1995. We conduct four strategies to refine the data: (1) Considering the generality, the language is restricted to English. (2) As for document types, we choose "article" and "review" in that they are strictly reviewed and judged by experts. (3) Categories are limited to "Business, Management, Economics," and other four categories "Sociology, Political Science, Social Sciences Interdisciplinary, and International Relations." "Business, Management, and Economics" focus more on firms' legitimacy. Legitimacy is considered to be an important subject in Political Science and Sociology (Suddaby et al., 2017). International Relations and Social Sciences Interdisciplinary may cover the legitimacy of a government or multinational firms. Thus, these four categories are included to cover a wide range of studies. We get 7,863 records. The "Plain Text" files of these records are then downloaded in the form of "Full Records and Cited References." (4) To ensure that all the articles cover the two evaluation objects (firm or government) and meet the evaluator's perspective, we manually review the titles and the abstracts of all these articles. Articles without mentioning firm or government legitimacy in the eyes of different evaluators are excluded. Meantime, the articles without abstract or cited references are deleted in this process. The final sample then consists of 1,203 articles that closely relate to our research objective.
Data processing techniques and tools. First, HistCite is employed to conduct the citation analysis. It is commonly used to assess the citation characteristics in the form of tables. It provides clear and useful informetric indicators: (1) Recs, refers to the number of articles; (2) (T)GCS, refers to "(total) global citation score, namely the (total) citation in WoS database; (3) (T)LCS, refers to "(total) local citation score," namely the (total) citation within the retrieval collection of the present study; (4) ALCS, refers to average (total) local citation score, namely the ratio of (total) local citation score to Recs (Garfield et al., 2006). Hence, we mainly use TLCS due to its close relation with our study, and we also list TGCS for reference. Besides, the historiographic map it offers is useful to locate important documents and find the history of document development. The historiographic map is further exported to Pajek to visualize the main path diagram.
Second, CiteSpace is employed to conduct the co-citation and collaboration analysis. It provides a convenient tool to display the main areas of a certain topic, and we can further analyze the main areas according to the reference derived from the visualization. The findings are visualized by VOSviewer and CiteSpace, respectively. The initial CiteSpace settings are displayed in Table 1. We also use CiteSpace to normalize the citations and remove duplicates.

Author Distribution
The top 20 authors with at least 47 TLCS are displayed in Table 2. Among all the 2,319 authors, Suchman MC is the most eminent author, ranking first in terms of TLCS with only one article, and he also enjoys the highest TGCS (6,584). Palazzo G is most productive and ranks second in light of TLCS, followed by Scherer AG. Agle BR and Mitchell RK rank fourth and fifth, respectively. Noticeably, the top five authors enjoy relatively higher TLCS and are more influential among the top 20. Of the top 20 authors, 11 authors come from US institutions; 6 from Canadian institutions; 2 from Swiss institutions; and 1 from China.
The author's cooperation network is formed by CiteSpace and visualized by VOSviewer for aesthetic purposes ( Figure  2). Different colors refer to different cooperation teams. The circle refers to the weight of the author. The distance between two nodes refers to the relatedness of the two authors. The values of network nodes, links, and density are 59, 16, and 0.0094, respectively. Roughly speaking, four teams are relatively major and have formed different levels of cooperation. Network #1 includes Scherer, Schneider, Palazzo, and Spence; Network #2 contains Castello, Bakker, Lozano, and Ettor; Network #3 consists of Marano, Berrone, Tashman, and Kostova; Network #4 includes Zavyalova, King, Deephouse, and Pfarrer.

Institution Distribution
Among 1,055 institutions, the top 20 institutions with at least 50 TLCS are listed in Table 3. Notably, the University of Lausanne possesses the highest TLCS (282), suggesting its great influence in this field. The University of Zurich takes second place with 230 LCS. The University of Pittsburgh is the third with 183 LCS, and it has the highest TGCS (5,992) and a higher ALCS (61). Of the top 20 institutions, 10 institutions are from the USA, followed by 5 Canadian universities, 2 Swiss universities. About 19 institutions belong to developed countries, and only City University of Hong Kong comes from a developing country.
The institution collaboration network is obtained by CiteSpace and visualized by VOSviewer ( Figure 3). The values for the nodes, links, and density are 206, 187, and 0.0089, respectively. Different colors refer to different cooperation teams. The circle refers to the weight of the institution. The label of the circle is proportional to the weight. The distance between two nodes refers to the relatedness of the two institutions. Copenhagen Business School and University Carlos III of Madrid are the cores that constitute the complex cooperative network. Copenhagen Business School ranks first with 18 records, and it's one of the top three business schools in Northern Europe. University Carlos III of Madrid takes second place and it's one of Spain's top universities, enjoying a high reputation in the fields of economics and business.

Country Distribution
Among 67 countries, we present the top 20 most cited countries with at least 13 TLCS in Table 4. The USA is a predominant country in light of TLCS (1,872) and Recs (381), followed by Canada (718) and Switzerland (363). Of the 20 countries, 19 countries belong to developed countries, and only 1 developing country (China) is on the list. Even though China ranks fifth in terms of TLCS, ALCS is far below Switzerland, Canada, and the USA.
The country collaboration network is obtained by CiteSpace and visualized through VOSviewer ( Figure 4). The values for the nodes, links, and density are 40, 168, and 0.2154, respectively. The color of the line between two nodes refers to the cooperation relationship of the two countries. The circle indicates the weight of the country. The label of the circle is proportional to the weight. The distance between the two countries suggests the relatedness of the two countries. The top three countries with extensive collaborations are the USA, England, and France. In general, the collaboration among countries in the evaluator's perspective is close, and cooperation between developed countries is closer.

Journal Distribution
The top 20 journals with at least 18 TLCS are displayed in Table 5, the publications of which account for 36% of the whole sample. Three journals are noteworthy in terms of TLCS: Academy of Management Review, Journal of Business Ethics, and Academy of Management Journal. Academy of Management Review is the most outstanding with both the highest TGCS (16,077) and the highest ALCS (40.04), suggesting the high quality and influence of this journal. Besides, the Journal of Business Ethics has contributed the most articles among the top 20, far ahead of other journals, which can be viewed as the main platform to publish articles in this area. Journals from the USA, Netherlands, and England dominate the list.

Article Citation Analysis
Research content. The top 30 articles with at least 24 LCS are outlined in Table 6. The most seminal article is written by Suchman (1995), which provides a basis for further study of legitimacy from the evaluator's perspective. Overall, the classical articles try to answer the following questions: theoretical basis, subjects of the evaluation, judgment content, judgment process, the influence of those evaluations, and firm's strategy to influence the evaluator's judgment. We analyze these questions according to the above-mentioned time division.
Theoretical basis evolution.
Evolution path. The citation paths of the top 30 highly-cited publications are visualized by HistCite ( Figure 5). The line represents the citation relationship between the nodes, and the nodes pointed to by the arrows indicate the cited documents. Overall, except for #56 and # 138, which are isolated nodes, the other nodes are connected. We further export the network file to Pajek to visualize the main path of studies from the evaluator's perspective ( Figure 6). Each node indicates a document, and the annotation on the right side of the node is the serial number in HistCite, author, and publication year of the document. The reference relationship is back and forth along the arrow direction, which is divided into five main paths. The citation network is originated in Suchman (1995), suggesting its pivotal role in the evaluator's perspective of legitimacy theory. Five main paths are as follows: (1)  Path 1 evolves from #35 to #574. It starts from the study on stakeholder identification, then shifts to reputation research and the difference between reputation and legitimacy, finally comes to the study on the legitimacy process of evaluators. Paper #35 (Mitchell et al., 1997) theoretically figures out the identification and salience of stakeholders. Paper #53 (Deephouse, 2000) focuses on media evaluators and presents the concept of media reputation, which can be viewed as a strategic resource to gain legitimacy and further raise the performance of commercial banks. Paper #107 (Deephouse & Carter, 2005) distinguishes legitimacy from reputation and further identifies the influencing factors of them. Paper #574 (Bitektine & Haack, 2015) develops a multilevel theory of the legitimacy process, notices the existence of "silenced" legitimacy and judgment suppressor factors, and explores competitive strategies due to the emergence of competing judgments. Path 2 evolves from #106 to #401, representing the change of firms' strategies from market strategy to the combined use of market strategy and political strategy. The contradictory institutional environment and the active response of firms are highlighted. Besides, firms' identity as political actors diverting scholars' attention to moral legitimacy. Paper #106 (Suddaby & Greenwood, 2005) mainly investigates the function of rhetorical strategy in legitimating profound institutional change. Paper #130 (Palazzo & Scherer, 2006) finds that with the rise of firms' new role as political actors, moral legitimacy becomes a new aspect for evaluators to judge, and captures an active communicative framework by CSR discourse rather than a passive response like compliance. Paper #401 (Du & Vieira, 2012) focuses on the CSR strategy and CSR communication tactics to meet the demands of different stakeholders in an environment of controversy. They find the combined use of different strategies to gain legitimacy.
Path 3 evolves from #64 to #369, shifting from theoretical research to empirical research on social legitimacy and from limited firm stakeholders to wide social actors. The article by Suchman (1995) influences the three articles in this path. Paper #64 (Lounsbury & Glynn, 2001) examines the role of cultural entrepreneurship in acquiring firm legitimacy from investors, competitors, and consumers. Paper #145 (Dacin et al., 2007) stresses the role of strategic alliances in legitimating firms and explores the influencing factors of such alliances. Paper #369 (Campbell et al., 2012) provides strong support for the hypothesis that CSR can demonstrate a firm's social commitment to improving its social legitimacy in host countries.
Path 4 evolves from #350 to #491 and explores firms' legitimacy in the eyes of the government by shifting from the western context to transitional economies because scholars believe firms in transitional economies are highly dependent on the national government. Paper #350 (Wang & Qian, 2011) explores the role of corporate philanthropy in gaining political legitimacy in the context of transitional economies and the firm heterogeneity (e.g., social expectation and political access) of conducting philanthropy in eliciting positive stakeholder response to gain better financial performance. Paper #491 (Marquis & Qian, 2014) mainly proposes that a firm's political dependency on the government exposes them to different legitimacy pressure, and such pressure influences the firm's likelihood to enact symbolic or substantive CSR.
Path 5 includes only one document by Hillman and Wan (2005). They concentrate on the determinants of MNEs' political strategy when MNEs confront pressures from both the corporation (internal legitimacy) and the host country (external legitimacy).

Thematic Evolution Analysis
Co-citation clusters of references are regarded to represent the distribution of thematic structure. The complete citation information of the 1,203 articles is used to extract the references and co-citation networks by using CiteSpace (Version: 5.8.R3). A final co-citation network with 424 nodes and 945 links is formed to be clustered. Two parameters are employed to depict the rationality of the clusters. Modularity measures the reasonability of the overall cluster structure. It falls within [0,1], and the overall cluster structure is believed to be significant if the value is above 0.3. Silhouette measures cluster effectiveness. The clusters are reasonable if the silhouette is above 0.5, and convincing if the value is above 0.7 (Chen et al., 2009). In our study, the modularity is 0.7479, and the silhouette is 0.9069, indicating the result of clusters is significant and convincing.
We get 13 major clusters (Figure 7). The color of the cluster relates to the year that two references are initially co-cited. Node circle refers to the citation frequency of reference. The larger the node circle is, the higher the frequency of the reference is. The lines between the nodes show the direct co-citation relationship, and the thickness indicates the strength of the co-citation relationship. The representative information of each cluster is illustrated in Table 7, including Cluster ID, size, silhouette, mean (year), and representative articles. Size refers to the number of references included in a cluster. The bigger the size is, the smaller the ID is. Silhouette indicates the homogeneity of the cluster with the max score close to or equal to 1, demonstrating the reliability of the results. The lowest silhouette score in our study is 0.821, suggesting the high reliability of each cluster. Mean (year) indicates the mean year of all the publication years within a certain cluster.
Thematic information is generally expressed by the cluster label. According to the above time division, we divide these clusters based on the mean year of every cluster into three phases accordingly.
Phase I (1995-2007). Nine clusters fall within this phase. Cluster #8 is a less correlated cluster and the earliest theme, concentrating more on the "institutional entrepreneur." Entrepreneurs should develop the ability to manage the environ-ment, identify stakeholders and their salience (Mitchell et al., 1997), and carry out symbolic management by utilizing verbal accounts to earn, maintain, or repair legitimacy (Elsbach, 1994).
After 2000, there are five clusters. Cluster #5 focuses on "organizational identity." Rhetorical strategy or cultural accounts are used to construct organizational identity when firms face institutional duality (Greenwood & Suddaby, 2006) or the fast-changing institutional environment (Greenwood et al., 2002). Cluster #10 focuses on "intangible assets." Environmental information disclosure, media reputation, cultural entrepreneurship, strategic alliance, and endorsement relationships are viewed as intangible assets for firms to gain legitimacy. Cluster #1 emphasizes "business ethics." To create a better business-society relation, firms adopt an isomorphism strategy and shoulder a new identity as "corporate citizenship" to meet the demands of social norms and expectations, to relieve the growing tension over corporate involvement in wider social life (Margolis & Walsh, 2003;Matten & Crane, 2005). Cluster #9 focuses on "network legitimacy." Political strategy, guanxi strategies, and interpersonal ties are mentioned to respond to external or internal isomorphism pressure. Cluster #6 focuses on "leadership." The leadership of senior managers is highlighted to perform identity management, like rhetoric (Suddaby & Greenwood, 2005) and identity crafting (Lounsbury & Glynn, 2001), to gain legitimacy.
After 2005, there are three clusters. Cluster #4 concentrates on "environmental disclosure" since firms confront issues concerning environmental legitimacy (Aerts & Cormier, 2009). The role of CSR and the differences between annual environmental disclosures and environmental press releases are discussed when firms confront pressures from diverse actors (Aguilera et al., 2007). Cluster #7 concentrates on "strategic alliance." Legitimation by strategic alliance and its influencing factors are examined (Dacin et al., 2007). Besides, strategic alliances with non-governmental organizations (NGOs) and nonprofit organizations (NPOs) are explored since the emergence of NGOs and NPOs. Cluster #12 mainly talks about "activists." The impacts of activists (like NGOs) on a firm's CSR implementation, firm investment plans and projects, are discussed by scholars.
Phase II (2008II ( -2012. This phase includes three closely correlated clusters. Cluster #2 discusses "CSR communication." The politicized concept of CSR is proposed (Scherer & Palazzo, 2011). Scholars further figure out the CSR difference among countries, and the difference between positivist and postpositivist CSR (Matten & Moon, 2008;Scherer & Palazzo, 2011).Cluster #3 focuses on "reputation." Reputation is regarded as a form of social judgment, which affects judgment process (Bitektine, 2011) and the firm's earnings and investor's reactions (Pfarrer et al., 2010). After 2010, Cluster #0 focuses on "signaling theory." Institutional complexity which causes multiple and competing demands, and institutional changes are emphasized in this phase, thus signaling theory is widely used to analyze governmental signaling which is an important mechanism of political influence, and a firm's response to the signaling (Marquis & Qian, 2014).
Phase III (2013III ( -2020. Cluster #11 falls within this cluster and focuses on legitimacy theory. To deal with threats to firms' social legitimacy, scholars combine legitimacy theory with signaling theory or economics theory, to assess the quality of various reports concerning sustainability, environment, or CSR, to assess the role of voluntary disclosure and symbolic or substantive disclosure, to assess the transparency or comparability of the reports (Boiral, 2013;Cho et al., 2015;Cormier & Magnan, 2015).
It can be seen from the timeline view that #0, #2, #3, #4, and #11 keep active in recent years and are relatively new research.

Intellectual Turning Points
The nodes of co-citation clusters connect different thematic information, which can be viewed as the intellectual turning points (Chen et al., 2009). Betweenness centrality is widely used to measure the number of shortest paths passing through a node; Thus, it acts as a bridge between two other nodes (Díez-Martín et al., 2021). The node with high betweenness centrality (above 0.1) will be a significant connector between two or more nodes (Chen et al., 2009). We find 14 publications with betweenness centrality above 0.1 (Table 8). These articles serve as the intellectual backbone of this field. The reference co-citation network is exported from CiteSpace to VOSviewer to visualize the map of cluster references (Figure 8). Connections between two nodes with high betweenness centrality are linked by the red lines.
Cluster #5 acts as the main intermediary to many other clusters due to the emergence of five nodes with high betweenness centrality in this cluster. Creed et al. (2002) present a framework considering both legitimating accounts and the construction of social identities. This paper connects cluster #5 to CSR communication (#2). Suddaby and Greenwood (2005) mainly discuss rhetoric as a strategy to legitimate profound institutional change. This paper connects cluster #5 to leadership (#6) and intangible assets (#10). Zimmerman and Zeitz (2002) explore legitimacy impact, legitimacy source, strategy to acquire legitimacy for new ventures, and put forward the concept of legitimacy threshold. This paper connects cluster #5 to business ethics (#1). Greenwood and Suddaby (2006) seek to explain institutional entrepreneurship as a strategy when firms confront internal and external legitimacy challenges. This paper connects cluster #5 to reputation (#3). Aldrich and Fiol (1994) put forward innovative ventures to handle the legitimacy pressure faced by new organizations in emerging industries. It connects cluster #5 to institutional entrepreneur (#8).
Cluster #2 is the next key cluster including three nodes with high centrality. Matten and Moon (2008) concentrate on the CSR difference between countries and raise the concept of implicit and explicit CSR. This paper links cluster #2 with signaling theory (#0) and environmental disclosure (#4). Scherer and Palazz (2011) propose the politicized concept of CSR. Scherer and Palazz (2007) distinguish positivist from postpositivist CSR. They both links this cluster with business ethics (#1).
Cluster #0 includes two nodes with high centrality. Marquis and Qian (2014) capture the heterogeneity of legitimacy pressure and propose a political strategy (CSR) to cope with such pressure. This paper connects to CSR communication (#2). Lamin and Zaheer (2012) investigate the effectiveness of a firm's strategy by figuring out the different responses from different evaluators and further figuring out the separate moralities scattered among the public.
Cluster #1 consists of two nodes with high betweenness centrality. Matten and Crane (2005) raise the concept of corporate citizenship and its components; Margolis and Walsh (2003) also notice the corporate social initiatives. These studies connect the present cluster to CSR communication (#2).
Cluster #3 has one node with high betweenness centrality: Bitektine (2011) proposes the social judgment formation model by exploring the difference between legitimacy, reputation, and status. It connects to signaling theory (#0). Cluster #9 includes one node with high betweenness centrality: Hillman and Wan (2005) introduce the political strategy for firms to handle multilevel legitimacy challenges in and outside the firms. It connects to activists (#12).

Discussion and Conclusions
This study presents a bibliometric analysis of organizational legitimacy from the perspective of evaluators between 1995 and 2020. Literature data is obtained from the WOS core collection. The current research status, the evolution of the evaluator's perspective, thematic change, and intellectual turning points are demonstrated. The results are as follows.
We illustrate the present research status.
(1) Publication distribution analysis suggests that research in this field has evolved to be an important topic progressively, and is still very active, especially after 2008. The rapid development of social media enables the public to express their evaluation online, which may lead to the rise of extensive evaluators and promotes the development of the evaluator's perspective after 2008.
(2) Suchman MC is the most cited author, and Palazzo G is the most productive author. Collaborations among authors are not close. (3) The University of Lausanne is the most cited institution, while Copenhagen Business School is the most productive institution. (4) The USA is the most cited and the most productive country, indicating the advanced The "macro" and the "micro" of legitimacy: Toward a multilevel theory of the legitimacy process (Bitektine & Haack, 2015) 12 39 250 Note. Articles are shown in chronological order in each stage. We identify the evolution of the evaluator's perspective. The results of the content evolution of this perspective are as follows. (1) Theoretical basis moves from a single discipline to a multidisciplinary field by introducing social psychological theory. (2) Subjects shift from the firm's  stakeholders to the government and the public, indicating the generalization trend of subjects; The heterogeneity of subjects is captured recently due to the emergence of competing judgment. (3) The politicized and socialized trend of evaluation. Judgment contents shift from market behavior to non-market behavior. (4) The judgment process shifts from visible evaluation to invisible evaluation ("silenced"). (5) Evaluation influence shifts from the source of evaluation (direct, indirect) to the effects of evaluation (positive, negative). (6) Firm's strategy is shifting from passive strategy (e.g., isomorphism) to active strategy (e.g., commitment to the natural environment), from market strategy to political strategy, and from single strategy to multiple strategies. Additionally, the citation evolution path analysis identifies five development paths for scholars to trace back important literature. The five paths  Matten & Moon, 2008 0.21 "Implicit" and "explicit" CSR: A conceptual framework for a comparative understanding of corporate social responsibility Scherer & Palazzo, 2011 0.12 The new political role of business in a globalized world: A review of a new perspective on CSR and its implication for the firm, governance, and democracy Scherer & Palazzo, 2007 0.11 Toward a political conception of corporate responsibility: Business and society seen from a habermasian perspective 3 Bitektine, 2011 0.14 Toward a theory of social judgments of organizations: The case of legitimacy, reputation, and status 5 Creed et al., 2002 0.17 Clothes make the person? The tailoring of legitimating accounts and social construction of identity Suddaby & Greenwood, 2005 0.12 Rhetorical strategies of legitimacy Zimmerman & Zeitz, 2002 0.12 Beyond survival: Achieving new venture growth by building legitimacy Greenwood & Suddaby, 2006 0.11 Institutional entrepreneurship in mature fields: The big five accounting firms Aldrich & Fiol, 1994 0.1 Fools rush in? The institutional context of industry creation 9 Hillman & Wan, 2005 0.1 The determinants of MNE subsidiaries' political strategies: Evidence of institutional duality divert our attention to "competitive strategy, combined strategy, social legitimacy, emerging market, and political strategy." The above shifts and the main paths may enlighten the follow-up scholars to explore the trend from the evaluator's perspective. The result of the thematic evolution analysis shows that there are 13 research areas: signaling theory, business ethics, CSR communication, reputation, environmental disclosure, organizational identity, leadership, strategic alliance, institutional entrepreneur, network legitimacy, intangible assets, legitimacy theory, and activist. Among them, signaling theory, CSR communication, reputation, environmental disclosure and legitimacy theory are more recent themes; they tend to handle threats to a firm's social legitimacy by revealing nonfinancial information as a complement to financial disclosure. In summary, the field is developing toward new debates.
Analysis of intellectual turning points reveals that 6 clusters including 14 articles act as the intermediaries for other clusters. The result provides a way for scholars to easily locate the path of a certain topic, which is convenient for them to discover the knowledge base and transitions of studies on a certain topic in this field. Following the study by Díez-Martín et al. (2021), for example, scholars interested in the political strategy of firms under internal and external legitimacy pressure may refer to the work by Hillman and Wan (2005); Scholars interested in the effectiveness of firm strategies and differently-perceived moral legitimacy can refer to the study by Lamin and Zaheer (2012).
The findings displayed in the present study are of significant interest for scholars to understand the evolution of organizational legitimacy from the perspective of evaluators. By providing the overall development characteristics and trends from the evaluator's perspective, this study brings some useful references to scholars. It's possible for scholars to better examine the intellectual structure and movements in the area of the evaluator's perspective. Meantime, the study provides the basis for not only scholars to select articles, authors, institutions, and countries, but also librarians to subscribe to journals.
Our study has several limitations. First, we manually select articles to make sure they are closely associated with our study, which may reduce the generality of the result. Second, the bibliometric analysis itself may have some limitations. Older articles tend to be more often cited than recent articles due to time advantage. Besides, the reason why an article is highly cited is complex, thus the quality and the relevance of a highly-cited article should be judged by additional complementary studies from readers. Moreover, further studies may explore the research by using other databases like Scopus.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.