The Intellectual Structure of Social and Sustainable Public Procurement Research: A Co ‐ Citation Analysis

: Public procurement has recently taken into account social and sustainable factors, increasing academic literature in this field. The aim of this paper is to map the intellectual structure of social and sustainable public procurement research by means of delimiting the scientific domain’s research traditions, its disciplinary composition, and influential research topics. Given that there is a literature gap in bibliometric studies applied to this field, we conducted a co ‐ citation analysis to identify the intellectual structure of this area of knowledge. Co ‐ citation analyses identify networks of interconnections and, consequently, detect the most (and the least) active research areas, being a good complement to traditional literature reviews. This article contributes to science development because it is the first paper to carry out a bibliometric analysis in the field of social and sustainable public procurement, as well as the first one to conduct a co ‐ citation analysis among public procurement research. Consequently, it is also the first article to detect which papers have become burst in this research field. The results show twelve different clusters of publications that were cited by researchers who wrote papers on social and sustainable public procurement. In other words, the sources of knowledge that scholars used as references are analysed, identifying papers that can be considered turning points, as well as those that became specially cited over a discrete period of time. Six different research trends were identified over the last decade in regard to social and sustainable public procurement research. The conclusions highlight the relevance of the findings, especially because they provide guidance to researchers when conducting literature reviews, given that the most significant journals and papers are identified.


Introduction
Public procurement is the procedure used by governments and administrations to purchase goods or services from professionals and companies. Public procurement represents a substantial part of the public investment in developed economies; consequently, it has become a pillar in the set of investments of these economies. Public authorities can make strategic use of this tool, in order to optimise the use of public funds and support the development of a more innovative, sustainable, and competitive economy. In this regard, the Organisation for Economic Co-operation and Development (OECD) considers public procurement to be a crucial pillar, given that it represents a sheer volume of spending, "in fostering public sector efficiency and establishing citizen's trust" [1]. For this reason, public procurement must be well-governed and their systems well-Over the last 11 years, from 2010 to 2020, there has been an increasing number of academic papers that deal with social and sustainable public procurement. Only 19.2 papers were published per year during the period 2010-2014 in the Social Science Citation Index (SSCI) database. This amount has dramatically increased over the years: 36.5 papers per year were published during 2015-2016 and 70.8 published papers in 2017-2020. This increasing number of publications on social and sustainable public procurement makes it complicated to track this developing research area.
When knowledge increases in any field, it must be somehow analysed. Social and sustainable public procurement research is not an exception and literature reviews in this field have been published [15,[30][31][32][33]. One of the main contributions of these reviews is that they establish the conceptual and theoretical framework. For this reason, they are of vital importance for the future development of this area. Nevertheless, some items have not been identified despite being key aspects to understand how the research field has evolved and its current situation. Thus, the following questions remain answered: what are, currently, the main fields of research on social and sustainable public procurement? By means of what papers has dissemination of the field of knowledge been possible? What papers attracted more attention from scholars, thus setting the research trends, and when did that happen? To put it another way, it is necessary to define the intellectual structure of this area of knowledge. Consequently, scholars would be able to identify the objectives of future research, so that the research they conduct may contribute to current areas of research and create new fields of study that can be included in this field of knowledge.
The bibliometric methodology allows the analysis of large amounts of data from hundreds of academic publications. It offers the possibility of evaluating quantitative data and inferring qualitative aspects. Nevertheless, bibliometric methods, such as the analysis of citations or co-citations, must be accompanied by a large knowledge of the field of study [34]. If this analysis is carried out properly, good insight into the subject studied can be achieved, including: clustering, intellectual turning points, burst papers detection, and so on.
There are two main ways to analyse academic research: quantitative and qualitative. The qualitative approach is particularly focused on expert knowledge on the subject of study and the review of existing literature, while the quantitative approach has to do with bibliometric methodologies. The qualitative methodology mainly needs the subjectivity of the person who performs it. It does not seem to be the best way to establish the intellectual structure of a field of knowledge, or at least the only one to be taken into account. The speed with which new works are published, the quantity of them, the task of selecting the most relevant ones, among others, can easily overwhelm the ability of researchers to process all of that information. Moreover, the results obtained may show or reveal some subjective influence of the researcher's point of view [35,36]. In an attempt to avoid this weakness, in this work, we opted for a quantitative approach, and the use of the methodology based on the analysis of citations (specifically, of co-citations) to build an approximation of the research activity in this field of study.
This article mainly aims to determine and offer a vision of the intellectual structure and dynamics of social and sustainable public procurement research. This implies delimiting the scientific domain's research traditions, their disciplinary composition, and influential research topics [37]. Whereas classical reviews only increase the conceptual understanding of a research area, this article contributes to the field by means of analysing its intellectual structure. Furthermore, we provide researchers with a basis for the future development of this field of knowledge by means of identifying the main research areas, the main contributions that have led to its dissemination, and the trends that have occurred, which have captivated the attention of academia.
The literature gap that we identified is the absence of a bibliometric analysis of sustainable and social public procurement, as well as a co-citation analysis of public procurement, according to the results obtained by means of research conducted in the Web of Science. Therefore, the contribution of this article to the field is threefold: (1) it is the first paper that carries out a bibliometric analysis among the academic literature that deals with social and sustainable public procurement; (2) it is the first paper that conducts a co-citation analysis in the field of public procurement research; and (3) it is the first article that detects which papers have become burst in the public procurement research, since this is a very specific feature of co-citation analysis.
The structure of this paper will be as follows: methodology will be analysed in Section 2. Section 3 will show and discuss the results of the co-citation analysis. This will include clusters definition, intellectual turning points, and the burst areas (or in other words, the most active areas of research). Section 4 ends with a conclusion and recommendations for further research.

Methodology
With the objective of obtaining some approach to the intellectual structure of social and sustainable public procurement, a bibliometric study has been carried out on papers that focused on this subject. This is one of the methods that helps scholars to understand the origin and development of a field of knowledge. There is no doubt that bibliometric analyses have become a good complement to the traditional literature review [35]. Bibliometric methodology aims to examine the publication performance of researchers and to disclose the structure and dynamics of science [38], assisting in the discovery of old and current fields of research, as well as suggesting new ones, and pointing out possible gaps [39]. A search in SSCI shows three papers that deal directly with bibliometric methods applied to the field of public procurement. These are studies on circular economy business [40], green public procurement [41], and sustainable public procurement policies for the promotion of scientific innovation [13]. The existence of only three bibliometric analyses is consistent with the fact that this research area is relatively young (under development). Nonetheless, a co-citation analysis was not carried out in these bibliometric papers.
Among the different methods of bibliometric citation analyses, we have used cocitation, since it is one of the most widely used and validated [38,39]. The use of the cocitation method is especially recommended when the aim of the research is focused on answering any of the following questions: what is the intellectual structure of the field? What are the central, peripheral, or bridging research documents in the field? How has diffusion of the field taken place? [38]. The analysis of bibliographic citations, as is well known, is based on the premise that if an author cites the work of another, there must be some kind of interest in doing so; and, a bibliographic citation indicates some type of relationship between the citing work and the cited work. A co-citation is defined by Small [42] as the frequency in which two works are cited together. Thus, two papers are co-cited if both are included among the bibliographic references of the same work. In this way, assuming that the co-cited articles will have related content, the similarity between publications is assessed on the basis of the amount of overlap in their bibliographic references. This type of co-citation analyses also offers the possibility of filtering outstanding papers in a scientific area. Although the citation count measures the relative influence of a document, through the analysis of co-citations, it is possible to detect interconnections between papers, to identify networks, and to find important changes in lines of thoughts and paradigms [38]. Co-citation frequency measures the level or degree of association between two documents and allows the discovery and identification of scientific disciplines. Therefore, co-citation analysis makes it possible to map the intellectual structure of a research field and to detect trends in the research area (by the authors who work on these topics and the nexus between them), and to discover first-line studies and highlight high-impact discoveries [39].
One of the most useful ways to quickly visualise and understand the intellectual structure of different types of study objects (papers, authors, funding entities, organizations, etc.) is the use of bibliometric mapping techniques [43]. Currently there is a variety of software programs [44] that allow, through the construction of bibliometric maps, the analysis of citations and co-citations, such as: CiteSpace, VOSviewer [45], SciMAT [46], and CitNetExplorer [47], BibExcel [48], Sci2Tool [49], among others. Each software programs has advantages and disadvantages [43] and, in general, allows the realization of different types of mapping visualization from temporal or geospatial network analysis to the detection of bursts [43], or a combination of all, or part of them, to achieve quantitative indicators and infer qualitative results.
CiteSpace software [50] has some advantages compared to the others, such as the identification of major turning points and growth themes, the identification of the main emerging research trends through the detection of bursts, etc. Along with the fact that CiteSpace has already been used in the field of business and economics to analyse the intellectual structure of different areas (ex: [51,52]), it has been the software used in this paper.

Data
All of the scientific journals included in the SSCI database have been selected in order to conduct the study, in view that SSCI is one the most frequently used databases to conduct bibliometric analysis in business and economics [38]. The search was carried out without any thematic limitation. Therefore, the results obtained belong to different research areas. In order to obtain the documents on social and sustainable public procurement, the following Boolean search strategy, in terms of titles, abstracts, and keywords (TS) has been used: (TS = "public* procurement*" OR TS = "government* purchas*" OR TS = "government* procurement*" OR TS = "public* purchas*" OR TS = "public* tender*") AND (TS = "soci* valu*" OR TS = "soci* account*" OR TS = "soci* effi*" OR TS = environment* OR TS=sustainab* OR TS = "soci* metri*" OR TS = "soci* impact*" OR TS = performan* OR TS = "stakeholder* theor*"OR TS = ethic*). The results obtained, segmenting the papers by means of Web of Science (WoS) categories, are the following: environmental sciences (130), economics (118), environmental studies (105), green sustainable science technology (91), management (91), and engineering environmental (53). Likewise, the main WoS research areas of the obtained papers were: business economics (218), environmental sciences ecology (169), science technology other topics (95), engineering (81), and public administration (65).
The time frame for the study (2010-November 16, 2020) and the thematic category parameters originate 452 citing papers containing 24,126 different references, which comprised the data sample of the analysis. Therefore, no citing article was excluded. Among the 232 journals in which the 452 papers are published, while some of them are highly specialised in their field, others have a multidisciplinary scope. Table 1 shows the number of papers on social and sustainable public procurement published by journals, including only those who published three or more papers. All sources of knowledge on social and sustainable public procurement (24,126 bibliographic references) found in the 452 citing papers obtained from SSCI were included in the bibliometric analysis. This does not mean that only academic journal articles were taken into account for the analysis. In fact, other types of documents (for instance, books, proceedings of conferences, etc.) were also included, since the analysis focuses on the cocited references of the 452 citing papers. Table 2 shows the parameters used to configure CiteSpace to perform the analysis (time slice, term source, node type, pruning, and selection criteria). CiteSpace also needs a careful definition of the source of the terms from which clusters are identified. In this study, among different possibilities, we chose the broadest option in SSCI by including the title, abstract, and keywords of the 452 papers found. In this way, it is possible to obtain a good range of terms for the analysis, but CiteSpace allows to complete it by including the terms that also appear in the 24,126 references cited (named in the program: "keywords plus (all)"). This improvement guarantees a greater and more complete source of terms that, in theory, should be well-related to the field of study to be analysed. An important decision to consider, to the greatest possible degree, is the node selection criteria in order to achieve a sufficiently cohesive network where clusters are different from each other, in spite of containing similar documents within them. CiteSpace gives different node selection criteria: g-index, Top N, Top N%, and threshold interpolation. In this study, the g-index was chosen. This criterion allows generating a network with the least number of small clusters and, therefore, improves the visualization of nodes and links [53].
The g-index improves the limitations of the h-index because it measures the overall citation performance of a set of articles [54] and gives more weight to the most cited articles, without the influence due to the total number of considered papers [55]. CiteSpace also introduces into the g-index a scaling factor k in order to regulate the total size of the network obtained. A k = 25 was chosen to develop the most suitable network for our set of analysed terms. In the Results section, the structural quality of the network obtained according to the modularity and silhouette indicators of CiteSpace is shown.

Results and Discussion
Over the last six years, social and sustainable public procurement research has had, as a whole, clear growth, showing pronounced take-off over the last four years (Figure 1). The growth was moderate between 2010 and 2014 (from 15 to 21 papers per year), with a considerable increase in 2015 and 2016 (36 and 37 papers per year respectively), and a clear upswing from 2017 to the present (between 54 and 79 publications per year). This research was mainly published in "environmental sciences", "economics", "environmental studies", "green sustainable science technology", "management", "engineering environmental", "public administration", "business", and "political issues" journals, according to the Web of Science categories. Considering the 11 year over the period between 2010 and 2020, the 452 published papers represent 41.1 publications per year. The publication of articles in journals included in the SSCI database, in other words, in high-impact journals, provides evidence that a field of knowledge was developing in social and sustainable public procurement-but up to what point? This is the aim of the third section of this paper.

Main Research Areas in Social and Sustainable Public Procurement
The main research areas in social and sustainable public procurement are shown in Table 3. The network is divided into 12 main clusters (from #1 to #12). Each cluster corresponds to a different thematic structure. The results obtained in the form of clustering show the impact of subsequent research co-citing. Take the example of cluster #1, which is formed by 63 publications. This means that many scholars used several of these 63 publications together as a source of knowledge. Consequently, cluster #1 is considered a thematic structure. Even though former researchers might have thought that their research was a contribution to one thematic structure, their papers may have eventually been used in alternative manners, which may derive into the inception of new avenues and, therefore, new clusters. All things considered, every single cluster corresponds to a line of research or thematic structure.   Analysis of public procurement procedures that include different social sustainability requirements.
In order to select the 12 clusters, the cluster silhouette value was used as a criterion. This value must range between 0.7 and 1.0 according to Chen et al. [50] and assesses the quality of a clustering configuration by means of cohesion and separation. Whereas cohesion refers to how similar an object is compared to its cluster, separation refers to how similar an object is to its own cluster in comparison with other clusters.
Furthermore, modularity Q assesses the quality of the overall network division. This value ranges from 0 to 1, according to Newman's method [57]. High modularity values suggest that clusters that have been created have clear boundaries. On the contrary, if this parameter is low, a bad-structured network has been carried out [58].
Every single created cluster (12 in total) has a silhouette value higher than 0.9. This means that there is good homogeneity between clusters. Modularity Q value is 0.7973, which means that the created network is reasonably divided into loosely coupled clusters. Figure 2 shows the sustainable and public procurement network. Nevertheless, the main limitation of co-citation analysis is that knowledge and experience of analysts are needed to identify clusters and interpret the core thematic structure. Consequently, evidence-based results are difficult to be distinguished from speculations and heuristics [38]. Both the common links between researchers of each cluster and citation have been analysed to summarise the essence of each cluster, as well as to enhance the robustness of the co-citation cluster labelling process Cluster #1 deals with the subject green public procurement, analysing whether its implementation is possible according to different criteria: size of municipalities, economic sector, manager attitudes, and so forth. Some articles analyse this issue in relation with the experience in some specific countries: Italian municipalities [17], Sweden [59,60], China [61], Malaysia [62], and Greece and Cyprus [63]. The authors use specific cases to carry out an analysis of the factors that promote or hinder green procurement. Conversely, Cheng et al. [15] address the subject on the basis of a comprehensive review of the literature from 2010 to 2016. Likewise, Xu et al. [64] put forth a set of indicators that allow to determine, with reliability, which product is more eco-friendly, and to, therefore, optimise public procurement processes.
Cluster #2 deals with private innovation, specifically how public procurement promotes and incentivises innovation in private companies. The key issue has to do with the incentive schemes created by the public sector in their procurement of goods and services procedures; ultimately, whether these procedures create (or not) frameworks that make it easy for the companies to innovate. A source for innovation is by contracting new products or services that need to be developed and provided by the private sector [65]. Guerzoni and Raiteri [66] show that research on this subject has focused on the supplyside policies: Research & Development (R&D) subsidies and tax credits, and less on the demand-side: public procurement. On the other hand, Rolfstam [67] considers that an institutional approach that takes into account this question from a wider perspective is necessary. This author recommends the implementation of institutional theories that, from a macroeconomic perspective, explains how the design of public institutional framework can produce certain results in economic agents; in this case, encouraging innovation in R&D. Likewise, Uyarra and Flanagan [68] propose an approach that pays attention to the great variety of public procurement policies and not just a few unrepresentative examples, since they consider that all public procurement can promote and incentivise business innovation, which can also be promoted by means of competition amongst cities and regions [69].
In Cluster #3, the subjects of sustainability related to public procurement and local SMEs development and, to a lesser extent, the reduction of the impact on environment, are discussed. In this cluster, the concept of sustainability is analysed in relation to local development, aiming to strengthen the SME business network by means of public procurement procedures. Administrations are expected to support the small and medium enterprises located in their territory; hence, special emphasis is put on the strategies of local governments. Walker and Brammer [70] and Walker and Preuss [71] study the way public procurement from small local suppliers in the United Kingdom has been used as an instrument for local and regional development, pointing out that there are significant differences among the various agencies and that the cost of products and services may act as a barrier. Seuring and Müller [30] make a mere review of the literature on the subject of sustainable supply chain. Furthermore, Preuss [72] studies this topic taking into account three aspects: purchases from small local suppliers, recruitment of voluntary organizations, and replacement of materials that are hazardous to persons and the environment.
Cluster #4 is focused on innovation in public procurement processes and the participation of SMEs in tenders. Whereas cluster #2 deals with the fostering innovation in supplier companies and cluster #3 has to do with the development of local SMEs, this one is focused on the internal procedures of public procurement in order to establish whether they are innovative and efficient. Thus, Georghiou et al. [73] analyse whether public procurement processes foster innovation or not; and Uyarra et al. [74] wonder whether public procurement procedures may act as barriers to innovation. On the other hand, Edler and Yeow [75] develop the concept of public procurement of innovation related to the intermediation between supply and demand in public procurement processes, and Knutsson and Thomasson [76] link innovation in the public procurement process to the strategic thinking of governments and administrations. From a different angle, Flynn et al. [77], Loader [78] and Reijonen et al. [79] investigate how public procurement procedures can hinder or reinforce the participation of small and medium enterprises in tenders.
Cluster #5 has to do with sustainable procurement and case studies. This cluster identifies and describes sustainable procurement procedures, particularly in terms of respect for the environment and cutting carbon emissions, on the basis of several cases. Brammer and Walker [80] and Meehan and Bride [81] highlight the relevance of the subject, with respect to the reduction of environmental impact, and analyse this practice by means of surveys to assess whether managers use this methodology or not; and Walker and Brammer [82] delve into the connection between this subject and e-procurement.
Many of the articles develop this question by means of practical cases: the healthcare sector and social services in Northern Ireland [83], the case of fifteen French public hospitals [14], the carbon reduction in the United Kingdom [84], the case of a distributor of garden products [85], and the implementation of sustainable procurement in the Netherlands [86]. Both Testa et al. [87] and Blome et al. [88] develop this question focusing on the topic of green procurement.
Cluster #6 deals with social and sustainable public procurement related to school feeding. In this cluster, the issue of sustainable procurement is analysed with regard to contracting suppliers for school meals, understanding the concept of sustainability in a broader sense, not only referring to environmental aspects, but also including social objectives. The European Commission [89] presents and develops the concept of green public contract in general, distinguishing it from social and sustainable public procurement: green contract is focused on the impact on the environment. Nevertheless, Sonnino [90], Wahlen et al. [91], Otsuki [92] and Bundy et al. [93] focus on the specific case of public procurement in school feeding. These policies are included in specific programs that have food improvement as their purpose, fostering purchases from local suppliers, and opting for environmentally sustainable food modalities.
The topic of Cluster #7 is public contracting or tendering by means of auctions. A comparison is made between this tendering procedure and other alternative systems, considering their advantages and disadvantages in terms of cost, commitment of the supplier companies and quality of the offer. Coviello and Mariniello [94] study how giving more publicity to an auction can increase the number of companies interested in bidding and, consequently, it contributes to the reduction of the tender cost. Likewise, Decarolis [95] describes this methodology in the case of Italian public administrations: the first auctions initially allow the reduction of prices but the poor performance that ensued because of the lack of commitment of the winning businesses led to a significant increase in final costs. Along this same line, Baldi et al. [96] consider that the institutions' quality and complexity improve the outcomes of awarding in open auctions, and Sánchez Solino and Gago de Santos [97] state that an open auction system is preferable to the negotiated procedures in public contracts. Ahsan and Rahman [16] analyse the subject of green public procurement in the healthcare sector of Australia.
Cluster #8 considers public governance and public procurement. This cluster analyses the structure of public organizations and institutions and the impact of this structure on public procurement procedures. In short, it is focused on the relation between governance, which is intimately linked to the design of public institutions, and public procurement. Glas et al. [98] analyse how a more or less centralised organisational structure in administrations results in differences in the design and implementation of the strategic objectives. On the other hand, Gelderman et al. [99] point out that there are significant differences in the objectives for procurement between politicians and purchasing managers, whereas Glas and Ebig [100] study public procurement by means of dividing tenders into lots. Patrucco et al. [101] analyse the systems to measure the local governments' performance in the procedures and public procurement objectives. Likewise, Patrucco et al. [102] conclude that there are opposed objectives in the aforementioned public procurement procedures. Finally, OECD [103] analyses in a comprehensive report the level of performance in the public sector of several countries, in particular members of the OECD.
The reviews of literature about circular economy are dealt with in Cluster #9. The concept of circular economy is developed by means of both a critical analysis and the definition and scope of this concept and by showing the contributions to the field of sustainability and environmental protection. Geissdoerfer et al. [31] review the bibliography related to the concept of circular economy and its relationship with the concept of sustainability, appreciating the ambiguity of this relationship. Likewise, Ghisellini et al. [32] confirm the great interest of circular economy as an alternative to growth and resource consumption, in particular in urban areas; and Lewandowsky [33] reviews literature in order to identify and classify the characteristics of circular economy depending on the structure of the business model. Finally, Alhola et al. [104] analyse the role of public procurement in the reinforcement of circular economy.
Cluster #10 has to do with case methodology; the work of Yin [105] the most representative of them, which analyses the use of the case study methodology as an alternative to the traditional experimental methods.
Cluster #11 deals with the subject of energy efficiency. In this cluster, suppliers are expected to offer goods and services that meet the requirements in terms of energy efficiency and water consumption; namely that they commit to saving energy and water. This cluster analyses to what extent is this requirement present in public procurement and the way to include it in the design of public procurement procedures. Singh et al. [106] analyse the question of public procurement related to the objective of increasing the efficient use of energy and the Department of Energy of the Office of Energy Efficiency and Renewable Energy U.S.
[107] studies the protocols to be followed in order to measure and verify energy and water saving The subject of public procurement and social sustainability is addressed in Cluster #12. The degree to which public procurement procedures include clauses to impose certain social objectives, such as the development of local communities, not only from an economic point of view, but also from a social and cultural perspective, is analysed. Bernal et al. [12] study by means of a Delphi analysis to what extent are social considerations taken into account in public tenders and Montalbán-Domingo et al. [108] analyse tender documents to identify the social sustainability objectives described in them. Finally, Loosemore [109] investigates whether building businesses in United Kingdom that take part in tenders have a positive impact on the communities where they build.
The mean publication year of the papers in each cluster is also shown in Table 3. On the one hand, clusters #12 (2017), #1, #8, and #9 (2016) are the clusters with a later mean publication year, which means that they are advancing at a quicker pace than the rest of them. On the other hand, clusters #3 (2008), #6, and #10 (2010) have an older mean. It does not seem to exist any pattern between the size of the cluster and the mean publication for either larger or more recent clusters. In the case of larger clusters, some of them are older (#3) or more recent (#1). Similarly, the smallest clusters are either older (#10) or more recent (#12).

Intellectual Turning Points in Social and Sustainable Public Procurement Research
Co-citation clusters are common thematic research structures. A node or dot is the representation in the graphic of every single publication. Those that connect different clusters are considered to be intellectual turning points [53]. Betweenness centrality quantifies the number of times a dot or node behaves as if it were a bridge along the shortest path between two other dots or nodes. By means of this methodology, the importance of a node when connecting others is valued. Therefore, a node with high levels of betweenness centrality can be considered as an essential connector between two or more nodes [53]. Betweenness centrality is correlated with future citations of the publication in the long term, according to a bibliometric approach [110].
According to social network theory, nodes with betweenness centrality higher than 0.10 can be considered as high central nodes. These nodes tend to be located on the paths that connect different clusters [53]. Turning points have helped the most to disseminate the field of research on social and sustainable procurement, acting as bridges between different clusters. These publications are connected to a large amount of research papers in the field. This implies that a great number of scholars consider these publications as an intellectual resource. Table 4 shows seven papers whose betweenness centrality is higher than 0.10 on social and sustainable public procurement. These papers can be considered to be the intellectual backbone of this research field. The average number of papers with high betweenness centrality (>0.10) per cluster is 0.58 per cluster. The research area that spread the most knowledge and that has the most intellectual turning points is cluster #3, with two papers with betweenness centrality higher than 0.10. Clusters #1, #2, #4, #5, and #7 have only 1 turning point among their notes. Lastly, the least connected research areas are clusters #6, #8, #9, #10, #11, and #12 given that they have no intellectual turning point. Journal of Cleaner Production The paper with the highest betweenness centrality is by Lember et al. [69]. Despite the fact that public procurement for innovation represents a policy tool on the demand side with high potential for the improvement of urban competitiveness, it has been scarcely studied. Evidences suggest that local and regional governments are getting more involved in the achievement of innovative solutions, but general knowledge about the ways in which they are involved and the effect of their participation, is still limited. Based on a study of cities in the northern part of the Baltic Sea, the authors revealed that public procurement that fosters innovative solutions had a positive impact on suppliers and that, in this sense, authorities might act as market creators as they promote the demand of more innovative products and services. The study also shows that, despite this, public procurement is not regarded as an inherent part of innovation policies. In this sense, there is a lack of conscience among the civil servants of these cities about the connection between procurement and innovation. Furthermore, local authorities are not inclined to take on the risks inherent to the encouragement of innovation through public procurement. Nevertheless, because of its positive effects, a greater development of public procurement policies adopting this approach is recommended.
The second paper with the highest betweenness centrality is by Georghiou et al. [73], where public procurement is regarded as an important instrument for the promotion of innovation and performance in private sector. The article aims to address the basis of public procurement policies to ascertain how administrations promote innovation. It establishes a wide taxonomy of public procurement policies that have appeared in the OECD countries. Then, they are compared with the companies' perceptions, according to a survey to suppliers in the public sector in the United Kingdom. It shows that these policies include the creation of framework conditions that allow to establish organizational frameworks, to identify the needs and to incentivise innovative solutions. Nonetheless, the results confirm that barriers to innovation faced by companies correspond, to a great extent, to the weaknesses in the design and implementation of these policies, despite the fact that they can also be explained by other factors attributable to companies such as, for instance, risk aversion.
Thirdly, Walker and Brammer [70] analyse the subject of sustainable procurement in the United Kingdom public sector. The analysis of quantitative and qualitative surveys reveals that there is a significant variation among public agencies with regard to the sustainable procurement practices. Thus, for instance, local authorities tend to purchase to small local suppliers. On the contrary, the healthcare sector is not particularly sensitive to this question, whereas in the case of the education sector, environmental questions aspects are taken into account in public procurement processes. The authors conclude that the main barrier for sustainable procurement is the cost. The government of the United Kingdom has the purpose of developing sustainable procurement policies and evidences suggest that progress is being made in this regard.
Fourthly, Ho et al. [111] examine procurement practices and policies in five Asian countries and Hong Kong, with the aim of elaborating a general description of their ecological procurement policies. The conclusion is that the level of governmental participation and public leadership in green procurement contribute significantly to ensure its efficiency and that updated information clearly improves the performance of this type of procurement.
Fifthly, Uttam and Le Lann Roos [59] discuss the way the building sector has adopted ecological public procurement to improve environmental performance. A description of the concept of green public procurement is expounded and its increasing importance is discussed. Administrations negotiate with contractors the conditions of the products and services they are purchasing, requiring them to comply with strict requirements of respect for the environment, thus ensuring that contractors get involved in the development of sustainable solutions. In this regard, the importance of the participation of suppliers in the design and implementation of these requirements is highlighted. This participation belongs to what authors define as "competitive dialogue procedure", a methodology that ensures the commitment of all the parties, particularly suppliers, in the final solutions.
The sixth paper with the highest betweenness centrality is Testa el al. [87]. Green public procurement is becoming one of the main thrust in environmental policies, both at the European level and in each of the member states. The paper analyses the pros and cons in the adoption of ecological procurement, based on the databases of three Italian regions' public authorities. Using an econometric approach, the aim is to test the following hypotheses: firstly, whether the existing conscience and knowledge on this topic and the practices, tools and regulations in force help public authorities to develop social and sustainable public procurement. Secondly, whether the support from external experts in the procurement function helps public authorities to develop these same strategies. Thirdly, whether the small size of public administration is an obstacle for their adoption. Fourthly, whether ISO 14001 (International Organization for Standardization) certification increases the possibility that the authorities develop the aforementioned strategies. The results show that the size of public administrations and the level of knowledge about the tools have a positive and significant effect on the probability that they adopt green public procurement policies.
Finally, Ahsan and Rahman [16] investigates the challenges in the implementation of a green public procurement system in the Australian public health system. By means of an extensive literature review, a framework for the implementation of this procurement typology, which includes different challenges classified in five categories, is developed. The problem is structured as a multi-criteria decision-making model based on the analytical hierarchy process. To that end, procurement executives in the healthcare sector were interviewed in order to obtain data and information. Some critical challenges were identified, among which were the lack of an adequate regulatory framework, the lukewarm support from senior managers, the absence of governmental ecological incentives for green procurement, and the lack of financial support. Some of the challenges that had also been identified, though less important, were those related to the ecological preferences in procurement, the understanding of the ecological policy, the supplier having the requested products and the collaboration with suppliers. The most critical challenges were mainly of a strategic nature, whereas the less important ones were operational. A clear understanding of the nature of these challenges in the making of decisions might help public health care institutions design strategies and policies for more successful green procurement.
With regard to centrality, the obtained results (Table 4) enable us to detect the journals where the most significant or influential articles in this field have appeared, taking into account that they are regarded as intellectual turning points: Journal of Cleaner Production (2), Urban Studies (1), Technological Forecasting and Social Change (1), Supply Chain Management: An International Journal (1), Natural Resources Forum (1), and Ecological Economics (1).

Burst Detection in Social and Sustainable Public Procurement
The most relevant indicators to deduce the impact of a field of study on scholars are the amount of quotations along with the number of publications, even though aspects, such as its influence, density, or evolution over time cannot be measured. Therefore, an in-depth analysis of the obtained clusters about publishing relations is needed. In this regard, citation burst is a more relevant indicator so as to identify the most active research areas during a specific period of time. CiteSpace identifies changes in a variable relative to others in the same population during time periods using the specific burst algorithm introduced by Kleinberg [112]. Accordingly, when it is detected that a publication receives an increasing amount of citations during a particular period of time, it can be said that a citation burst has taken place. In a brief space of time this paper has attracted great attention from other scholars. Consequently, if a cluster has several burst papers, it can be considered as an emergent and active research area [58].
The results of the burst detection analysis are shown in Table 5. This table also identifies the 26 papers with the largest citation bursts within the social and sustainable public procurement field from 2010 to 2020. The 26 papers were found to have citation bursts by the Kleinberg algorithm [112]. The rest of publications did not show enough bursts (See Appendix A). The analysis is right censoring for burst periods in 2020, therefore, we do not know the end date of the burst periods for papers that are currently burst. Sustainable procurement and case studies (cluster #5) is the area with the most burst papers. Eight papers in this area show strong citation bursts. The second most burstful cluster with five citation burst papers is #1, which deals with green public procurement. Thirdly, cluster #4 has to do with innovation in public procurement processes and has four burst papers. The average number of burst papers per cluster is 2.17. Nevertheless, only 6 clusters out of 12 have burst papers.  being the most prolific year in this regard. The last burst paper was written in 2016, therefore, no paper written from 2017 has become burst yet. This issue can be considered a paradox, given that the number of citing papers has considerably increased since 2017 (Figure 1). The average interval between the publication of burst papers and its maximum interest is 3.00 years. Whereas no paper became burst the same year in which it was published, one paper did it the next year after its publication [79]. Uyarra et al. [74] records the longest period of time in which a paper has been burst, namely 4 years.
Burst paper detection reveals research trends in a field of knowledge [113]. In this regard, there have been six trends in the field of social and sustainable public procurement research. Table 6 shows the six trends classified by cluster and indicates the number or burst papers, the year when this trend started (Min (year)) and finished (Max (year)), the mean year, the mean strength value, the year when that trend started (Min (begin)) and the year that the trend finished (Max (end)). The analysis of Table 6 shows that the first trend encompasses 4 papers of cluster #3, which deals with local SME Development. The publication of burst papers started in 2008 and finished in 2009. Nonetheless, they became burst from 2012 through to 2014. Just one year after the inception in the writing of the first trend, the writing of the second one started within cluster #10. This second trend comprises one single document [105] that was written in 2009 and establishes the methodology of case studies. This paper became burst in 2013 and 2014. The same year that these two trends vanished, in 2014, the third trend became burst. This third trend encompasses three papers that were published from 2011 (two years after the end of the writing of the first two trends) to 2015 and deals with private innovation (cluster #2). These papers were burst through to 2018. One year after the inception of the third trend, the fourth and the fifth one appeared (in 2015), which are the largest trends in terms of publications. On the one hand, the fourth trend was written in 2011 and 2012 and is formed by eight papers that became burst from 2015 to 2017. This is the largest trends in number of publications and deals with sustainable procurement and case studies (cluster #5). On the other hand, the second largest trend occurred within cluster #1 where five publications written from 2013 to 2016 became burst from 2015 onwards. In actual fact, this fifth trend that deals with green public procurement is still burst in 2020. Finally, the sixth trend was written in the same period of time as the fifth one (from 2013 to 2016) and became bust from 2016 and was still burst when this paper was written (2020). The sixth trend deals with Innovation in public procurement processes (cluster #4) and is formed by four papers.
All things considered, the second trend (cluster #10) shows that academics initially became very interested in achieving eminently theoretical understanding in the case methodology [105] and later on this method was applied by means of different case studies in the fourth trend (cluster #5), which is the largest one in terms of number of papers and the one that has the highest strength mean. For the time being, the fourth and the fifth trends are still alive and are the ones that have the highest longevity, namely the fifth trend that deals with green public procurement has been burst over the last 6 years, and the sixth one that deals with innovation in public procurement processes has been burst over the last 5 years. As a matter of fact, three out of the five burst papers, of which the fifth trend is composed (cluster #1), are still so in 2020.
A certain degree of continuity can be observed in the first trend (cluster #3), the third trend (cluster #2) and the sixth trend (cluster #4). In this regard, cluster #3 refers to the impact of public procurement in the development of small and medium local companies. These works analyse to what extent public procurement contributes to strengthening the local business sector, composed in its vast majority of small and medium size companies. These works-four in total-were written in 2008 and 2009. The three burst papers that are part of cluster #2 develop this concept, focusing on the subject of private innovation. These works were written during a period of four years (from 2011 to 2015), after the burst papers of cluster #3 were written. This research is focused on finding out the extent to which public procurement strategies and policies foster innovation in the small and medium suppliers when they are encouraged to invest in R&D. In fact, local development of SMEs makes no sense if it is not associated to promoting innovation. Otherwise, public procurement strategies might benefit inefficient companies and even keep artificially active companies that, under market conditions, would not be sustainable. For this reason, not only is the support to SMES emphasised, but, in particular, the support to the companies that invest, innovate, and create added value. The latter question allows the introduction of the issue of innovation in public procurement processes. Cluster #4 delves into this subject. The four burst papers included in this cluster, and that were written from 2013 to 2016, analyse the public procurement processes to ascertain whether these processes are efficient and innovative, or if, on the contrary, they are cumbersome and inoperative. The connection between this question and the promotion of innovation in SMEs is evident: public procurement processes can promote innovation in SMEs only provided that these processes, in turn, include systems and methodologies that can be considered innovative; thus avoiding contradictions between the objective of promoting innovation in suppliers, and the public procurement processes, by means of which this objective is intended to be achieved. Out of the four burst papers of Cluster #4, two continue to be so nowadays. Table 6 shows the Annual average of WoS citations received per paper. This ratio has been calculated by means of summing the total number of citations received by the papers that were burst in each cluster, according to Web of Science. This number has been divided by the number of papers that were burst in the cluster. The resulting number has been divided again by the average number of years since the papers were published. Therefore, this bibliometric variable shows the average number of citations that the burst papers in the same cluster received per paper and per year, in other words, the annual citations received per paper for each trend. The results suggest that the first trend, Cluster #3 that deals with Local SME development is the most cited trend followed by the last trend, Cluster #4 that analyses Innovation in Public Procurement Processes. The second trend (Cluster #10) is formed only by a single burst document and it is a book that is not indexed in Web of Science. For this reason, this variable has not been able to be calculated for cluster #10.
With regard to burst detection, the obtained results (Table 6)

Conclusions
The intellectual structure of social and sustainable public procurement research has been mapped by means of co-citation bibliometric analysis. The findings of this paper are quantitatively rigorous and contribute to further development of this field or research in the following ways.
Firstly, the main research areas within the field of social and sustainable public procurement research have been delineated: green public procurement; private innovation; local SMEs development; innovation in public procurement processes; sustainable procurement and case studies; school feeding; auctions; public governance; circular economy; case methodology; energy efficiency; social sustainability Secondly, the works that can be regarded as the intellectual structure of research on social and sustainable public procurement have been detected. By means of mapping these papers the paths and connections through which social and sustainable public procurement has disseminated has been revealed. The turning points are the foundations of the knowledge of the area of research. When they are identified, academia can gain a quicker understanding of the field, such as where to start their research. By way of example, green public procurement researchers should start their literature review by means of reading Uttam and Le Lann Roos [59]. Consequently, these works constitute the major sources of information in this particular field. Hence, future research in this field of knowledge should view them as the starting point, using them as the basis of their knowledge. following this procedure, in particular during the early stages in which the conceptual framework of the research is developed, could save a considerable amount of time, since they use the engines of knowledge of the field as references.
Thirdly, we identified the works that have drawn remarkable attention from the scientific community for a discrete period of time. A picture of the changing status of the social and sustainable public procurement literature is obtained by means of the identification of burst papers. Having this image is vital because researchers are provided with the guidance they need for the design of new research through which the detected trends are developed and also for the creation of new fields on the basis of these trends.
Out of the six trends, we can see a natural development among five of them. In this regard, the burst paper included in Cluster #10 [105] stablishes the case methodology so as to allow a later trend (Cluster #5) to apply this methodology by means of different sustainable procurement case studies. Similarly, the trends that include burst papers that deals with SME local development (Cluster #3), Private Innovation (Cluster #1) and Innovation in public procurement processes (Cluster #4) are written and become burst sequentially, although burst papers of Clusters #2 and #4 slightly overlap. For the time being, the trend that deals with green public procurement (Cluster #1), and innovation in public procurement processes (Cluster #4) are still burst, or alive, given that five out of nine papers that became burst are still so in 2020.
This means that there has been no new burst paper since 2017, even though the mean time from publication to maximum interest for a burst paper is 3.00 years. This can be considered a paradox considering that the number of citing papers on social and sustainable public procurement field (Figure 1) written from 2017 to 2020 is higher (283 citing papers) than those written from 2010 to 2016 (169 documents). Future research should address this issue by means of repeating the same co-cites analysis in a few years, to detect which burst papers will eventually become burst. It could happen that some of the papers that will become burst in the future have already been published.
Fourthly, the scientific journals with the most significant contribution to this area of knowledge have been studied. In fact, 232 different academic journals published the 452 citing papers. The journals with over five papers (Table 1) (7), and Amfiteatru Economic (6).
Nonetheless, it can be deduced that the papers with a greater amount of intellectual turning points (Table 4) and burst papers ( Table 5) were published in Journal of Cleaner Production, and at a lesser level, in Ecological Economics, Supply Chain Management: An International Journal, Technological Forecasting and Social Change, and Urban Studies. Having identified which journals published the most influential papers on social and sustainable public procurement will be of great help for future researchers, since they will need to spend less time on conducting the literature review.
Fifthly, by means of a bibliometric analysis, it is possible to give order to a huge amount of data; good quantitative and approximate qualitative evaluations can be extracted from data meaningfully ordered. It has some limitations though, which have been recurrent for decades [114]. To interpret the results that have been obtained, it is essential to have a high level of commitment and, consequently, subjective participation of those conducting the analysis. Therefore, bibliometric techniques need to be complemented by the intellectual refinement that requires, for instance, a thorough review of the bibliography on this subject, its discussion, and/or synthesis among researchers [38,115,116]. The researcher has to make some technical decisions with regard to both the chosen database's parameters and the analytical software in order to obtain a reliable set of indicators that could be used to reach some kind of reasonable conclusions (e.g., restrictions of language, journals, period of time, normalization of cluster labelling, verification of final data to debug errors, etc.) [115]. An additional problem has to do with the fact that, on account of the delay in which citations are done, older publications may have more presence than the newer ones [117] or the consideration or not of self-citations [118,119].
Sixthly, we recognise that the use of a single database (SSCI) is a limitation of the research. In spite of the fact that SSCI is well recognised by social sciences bibliometric analysis and, therefore, the results offered are well supported, many are the databases that could have been taken as a sample, such as Scopus or Dimensions [120]. Future lines of research could include either more databases or data offered by techniques applied to the analysis of scientific publication and activity on the internet, such as Webometrics or Altmetrics [121][122][123].
Seventhly, most of the studies and analyses of this topic are empirical articles about different countries, municipalities or economic sectors. Since these studies deal with public policies, it could be interesting to address them from a political perspective. In fact, there is not a cluster in which the topic is addressed from a political perspective, thus confirming the existence of a literature gap. For instance, within the modern political theory, the neo-institutionalist approaches would be highly suitable. Certainly, these approaches have proved to be very useful in different disciplines, namely sociology, economics, political science, international relations and historical analysis. Nonetheless, they only are referred to in a single article about sustainable and social public procurement [67]. It should be taken into account the fact that these kind of approaches are common in studies on public policies and also have been used in some articles about public procurement [124][125][126]. Nevertheless, they could be more widely used in order to relate the empirical side to the theoretical explanation of the link among sustainable public procurement specific policies, the organization of institutions and the performance of the involved agents. The nodal point of these neo-institutionalist approaches is that they can understand the agents' behaviour from the institutional framework in which they operate. This premise would allow to identify the main reasons for success and failure in the design and implementation of sustainable public procurement procedures, as well as the development of comparative studies. In this sense, future lines of research could focus on the analysis of the differences in sustainable public procurement among the countries with models more inclined towards markets self-regulation, and the countries with economic models based on institutionalised markets. Other approaches commonly used in certain social sciences that could also be applied to the analysis of sustainable public procurement are public choice, rational choice, or games theory, among others. The information has been already compiled and there is an abundance of empirical evidence. Starting with all of this information, it would be appropriate to give them a meaning and interpret them in future research, using the available theoretical models in the different academic disciplines and, in particular, in politic science.