Executive Summary

_{Author: Carlos Alfredo Salas Macías — Thesis director:}

^{Summary generated by the ResumeAI tool, an artificial intelligence solution developed by the Pecege Institute, focused on synthesis and writing.}

This research conducts a bibliometric and thematic analysis of scientific production on carbon markets from 1994 to 2024, to map the field’s evolution, identify trends, and reveal the structure of international collaboration. The study answers questions about the most relevant publication sources, collaboration patterns between countries, main research topics, and their temporal evolution and distribution. The methodology seeks to provide a quantitative overview of the state of the art, identifying thematic gaps and guiding future work in the fight against climate change.

Global warming is a pressing challenge, and carbon markets are a central mechanism to mitigate its effects by incentivizing the reduction of greenhouse gases (World Bank, 2021). However, their implementation faces challenges such as lack of consensus in designs, carbon price volatility, and the need for greater integration between economic and environmental policies (Kollmuss et al., 2008). Such problems justify a systematic analysis of existing research to identify trends and gaps that can strengthen scientific knowledge and public policies.

The scientific literature on carbon markets has grown exponentially. Annual production, stable at approximately eight publications per year between 2000 and 2010, increased from 2016 onwards, driven by the Paris Agreement (Zhang et al., 2022). The number of annual citations rose from less than 10 in 2000 to over 1,000 in 2014 (Yue et al., 2020). Recently, the annual growth rate of papers on carbon and financial markets has been approximately 14.65% per year, with a contribution that grew by more than 80% in 2020 alone (Dawar et al., 2024).

This expansion, however, has resulted in a fragmentation of knowledge. Topics such as equity in the distribution of benefits, the integration of voluntary and regulated markets, and the impact of technological innovations have not been sufficiently explored (Tumushabe et al., 2023; Zhang et al., 2022). This situation highlights the need for critical analysis to identify research gaps and guide future efforts. The analysis of scientific publications is a tool for understanding the evolution of knowledge, and this study employs a methodological combination for this purpose.

The present research contributes to the field by generating a structured database and robust methodological tools, promoting transparency and replicability (Aria & Cuccurullo, 2017). In this way, it seeks to offer inputs to the scientific community and decision-makers, strengthening global efforts to address climate change.

This study adopted a systematic methodology to analyze trends, thematic structure, and collaborations in carbon market research. The stages included data collection, descriptive analysis, topic modeling with Latent Dirichlet Allocation (LDA), and visualization using the HJ-Biplot technique. Data collection followed the PRISMA protocol (Page et al., 2021), using the Scopus and Web of Science (WoS) databases, recognized for their broad multidisciplinary coverage (Pranckutė, 2021; Chadegani et al., 2013).

The data collection procedure was divided into two phases. In the first, systematic searches were conducted in each database with terms such as “carbon markets”, “carbon trading”, and “carbon pricing”. The results were refined to include scientific articles and reviews, with the removal of duplicates and records without abstracts. In the second phase, the data were integrated with the R package ‘Bibliometrix’ and Excel, following an adapted methodology (Caputo & Kargina, 2022). Descriptive analysis calculated indicators such as the number of documents, the annual growth rate, and citations. A network analysis explored international collaborations, using the Walktrap algorithm (Pons & Latapy, 2006) to identify clusters, visualized with the Kamada-Kawai algorithm (Kamada & Kawai, 1989).

Topic modeling was performed with the unsupervised learning algorithm LDA (Blei et al., 2003), effective in identifying thematic patterns in large corpora (Griffiths & Steyvers, 2004). The process included text preprocessing (lowercase, removal of punctuation, numbers, and stop-words), model construction, and topic labeling, executed with LDAShiny (de la Hoz-M et al., 2021). To determine the optimal number of topics (k), models were generated with k ranging from 4 to 40, and coherence was evaluated with the CV metric (Röder et al., 2015), based on the hypothesis that words with similar meanings co-occur in similar contexts (Harris, 1954).

The visualization of topics and their relationships was performed with LDAvis (Sievert & Shirley, 2015), which maps the similarity between topics, and the HJ-Biplot technique (Galindo, 1986). HJ-Biplot is an evolution of traditional biplots (Gabriel, 1971) that optimizes the representation of rows (individuals) and columns (variables) in the same space, being a powerful tool for interpreting complex data structures. The technique has been applied to improve the analysis of the LDA model in several studies (Montes-Escobar et al., 2023; Abril-Spin et al., 2024). The interpretation of the HJ-Biplot diagram allows for a robust analysis of the relationships between topics, countries, journals, and temporal trends.

The statistical analysis of the set of 10,552 articles published between 1994 and 2024 reveals a dynamic research field. The documents come from 10,552 journals and have 18,319 authors. The collaboration index of 3.73 indicates strong co-authorship, although 1,315 works are single-authored. Scientific production showed exponential growth, especially from 2008 onwards. Until 2007, annual publications were less than 100, but from 2015 onwards, production accelerated, reaching a peak of 2,287 articles in 2024.

The journal Journal of Cleaner Production is the most prolific, with 362 publications and 14,103 citations. Other influential journals include Energy Policy (11,586 citations) and Climate Policy (6,151 citations). The diversity of journals, from broad-scope ones like Science of the Total Environment to specialized ones like Carbon Management, evidences the interdisciplinary nature of the research, which interconnects economics, politics, and technology. The geographical distribution of scientific production is concentrated in a few countries. China leads with 3,662 publications, followed by the United States (1,183) and Australia (614).

The analysis of international collaboration reveals that China, although a leader in volume, conducts most of its research (3,049 articles) independently. In contrast, the United States and European countries such as the United Kingdom, Germany, and France show a high rate of multi-country publications (MCP), indicating strong international cooperation. The collaboration network positions the United States as the global hub, with the highest levels of brokerage and connectivity. China, the United Kingdom, India, and Australia also act as key brokers, forming an interconnected blue cluster. A second yellow cluster, of European countries, demonstrates strong regional collaboration. The prominence of these countries, major emitters of greenhouse gases (IPCC, 2022), explains the research focus on carbon policies and pricing, although their heterogeneous approaches reflect different governance systems.

LDA topic modeling identified 32 topics as the optimal structure. The temporal evolution of these topics revealed clear trends. Emerging topics include China’s regional carbon trading (t_6), supply chain management (t_3), market dynamics and volatility (t_22), green technology investment (t_20), and predictive carbon pricing modeling (t_29). This increase suggests a growing focus on economic, financial, and technological solutions. On the other hand, declining topics include carbon project development (t_1), such as those under the Clean Development Mechanism (CDM), forest carbon estimation (t_4), agricultural soil carbon sequestration (t_12), and voluntary carbon markets (t_16). This decline indicates a maturation of the field, transitioning from initial concepts to the analysis of more complex, large-scale regulated systems, such as Emissions Trading Systems (ETS).

The visualization of the thematic structure using multidimensional scaling (MDS) revealed a division in the research. The first principal component (PC1) separates public policy and governance studies (regulation, taxation) from those focused on ecological and technological issues (carbon capture, renewable energy). The second component (PC2) differentiates high-level economic and financial research (pricing, investment) from operational and applied studies (logistics, supply chain optimization). This structure highlights the tension between regulatory strategies and ecological processes, and between macroeconomic analyses and implementation challenges. The HJ-Biplot analysis confirmed this specialization, showing that policy and economics journals cluster around governance themes, while technical and energy periodicals concentrate on modeling and technology.

The distribution of topics over time, visualized by the HJ-Biplot, organized the research into three phases. The initial phase (1994-2010) focused on foundational topics such as emission reduction and international agreements. An intermediate phase consolidated these concepts, while the recent phase (2011-2024) concentrated on applied topics such as the Chinese market, carbon finance, renewable energy, and predictive modeling. The country analysis also revealed specializations: a European cluster focused on policies and governance; an Asian cluster (led by China) focused on carbon trading and energy transition; and a third diversified cluster (USA, Canada, Brazil, Australia) with an emphasis on innovation, green technology, and mitigation strategies.

The thematic fragmentation identified between the policy/economy approaches and the technology/ecology ones is a significant finding. Although specialization is necessary, this disconnection challenges the development of effective carbon markets, which operate at the intersection of ecological, economic, and social systems. The need for integrated approaches, connecting governance, economy, technology, and ecology, emerges as a central conclusion. Overcoming this fragmentation is crucial for designing markets that are efficient, effective, and equitable.

This bibliometric study has mapped the evolution and structure of research on carbon markets from 1994 to 2024. The results show exponential growth in publications and thematic maturation, transitioning from ecological studies to applied approaches in economics, technology, and finance. The analysis also reveals a disciplinary specialization that fragments knowledge, with policy journals focusing on governance and technical periodicals on modeling. This fragmentation hinders the adoption of holistic approaches, which are essential for the complexity of carbon markets.

It is concluded that the objective was achieved: the evolution of research on carbon markets was demonstrated, identifying a transition from ecological to economic-financial themes, a growing geographical specialization led by China and the USA, and a disciplinary fragmentation that points to the critical need for greater interdisciplinary integration for the development of effective climate policies. Emerging areas, such as market integration, financial modeling, and behavioral economics, along with the need to strengthen North-South and South-South research networks, represent the frontiers for future research seeking to advance climate change mitigation through robust and equitable market mechanisms.

References:
^{April-Espin, E., Montes-Escobar, K., de la Hoz-M, J., Garcés-Fiallos, F. R., & Salas-Macias, C. A. (2024). Exploring global trends in scientific research on Rubus glaucus Benth.: A comprehensive analysis integrating bibliometrics, LDA, and HJ-Biplot. Journal of Berry Research, 14(4), 287–300. https://doi. org/10.1177/18785093241284955
Aria, M., & Cuccurullo, C. (2017). Bibliometrix: An R tool for comprehensive scientific mapping analysis. Journal of Informetrics, 11(4). https://doi. org/10.1016/j. joi.2017.08.007
Blei, D., Ng, A., & Jordan, M. (2003). Latent Dirichlet Allocation. Journal of Machine Learning Research, 3, 993–1022. https://doi. org/10.1016/b978-0-12-411519-4.00006-9
Caputo, A., & Kargina, M. (2022). A user-friendly method for merging data from Scopus and Web of Science during bibliometric analysis. Journal of Marketing Analytics, 10(1). https://doi. org/10.1057/s41270-021-00142-7
de la Hoz-M, J., Fernández-Gómez, M. J., & Mendes, S. (2021). Ldashiny: An R package for exploratory scientific literature review based on a Bayesian probabilistic model and machine learning tools. Mathematics, 9(14). https://doi. org/10.3390/math9141671
Dawar, G., Nagariya, R., Bhatia, S., Dhingra, D., Agrawal, M., & Dhaundiyal, P. (2024). Can financial markets help achieve carbon objectives? Evidence from a systematic literature review, bibliometric analysis, and topic modeling. Sustainability Accounting, Management and Policy Journal, ahead-of-print(ahead-of-print). https://doi. org/10.1108/SAMPJ-05-2023-0319
Freeman, L. C. (1977). A set of measures of centrality based on betweenness. Sociometry, 40(1). https://doi. org/10.2307/3033543
Freeman, L. C. (1978). Centrality in social networks: Conceptual clarification. Social Networks, 1(3). https://doi. org/10.1016/0378-8733(78)90021-7
Gabriel, K. R. (1971). The biplot graphic display of matrices with application to principal component analysis. Biometrika, 58(3). https://doi. org/10.1093/biomet/58.3.453
Galindo, M. P. (1986). An alternative for simultaneous representation: HJ-Biplot. Qüestiió, 10(1), 13-23.
Geman, S., & Geman, D. (1984). Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images. IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI-6(6). https://doi. org/10.1109/TPAMI.1984.4767596
Griffiths, T. L., & Steyvers, M. (2004). Finding scientific topics. Proceedings of the National Academy of Sciences of the United States of America, 101(SUPPL. 1). https://doi. org/10.1073/pnas.0307752101
Grün, B., & Hornik, K. (2011). Topicmodels: An R package for fitting topic models. Journal of Statistical Software, 40(13). https://doi. org/10.18637/jss. v040. i13
Harris, Z. S. (1954). Distributional structure. WORD, 10(2–3). https://doi. org/10.1080/00437956.1954.11659520
IPCC. (2022). Climate Change 2022: Mitigation of Climate Change. Contribution of Working Group III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. In Climate Change 2022 – Mitigation of Climate Change.
Kamada, T., & Kawai, S. (1989). An algorithm for drawing general undirected graphs. Information Processing Letters, 31(1). https://doi. org/10.1016/0020-0190(89)90102-6
Kollmuss, A., Zink, H., & Polycarp, C. (2008). Making sense of the voluntary carbon market: A comparison of carbon offset standards. Sustainable Development, 51(2).
Montes-Escobar, K., de la Hoz-M, J., Barreiro-Linzán, M. D., Fonseca-Restrepo, C., Lapo-Palacios, M. a., Verduga-Alcívar, D. A., & Salas-Macias, C. A. (2023). Trends in agroforestry research from 1993 to 2022: A topic modeling approach using Latent Dirichlet Allocation and HJ-Biplot. Mathematics, 11(10). https://doi. org/10.3390/math11102250
Montes-Escobar, K., Vicente-Villardon, J. L., de la Hoz-M, J., Useche-Castro, L. M., Alarcón Cano, D. F., & Siteneski, A. (2021). Frequency of studies on neuroendocrine tumors: Use of statistical methods of Latent Dirichlet Allocation and HJ-Biplot. Mathematics, 9(18). https://doi. org/10.3390/math9182281
Page, L., Brin, S., Motwani, R., & Winograd, T. (1998). The PageRank citation ranking: Bringing order to the web. World Wide Web Internet And Web Information Systems, 54(1999–66). https://doi. org/10.1.1.31.1768
Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., Chou, R., Glanville, J., Grimshaw, J. M., Hróbjartsson, A., Lalu, M. M., Li, T., Loder, E. W., Mayo-Wilson, E., McDonald, S., … Moher, D. (2021). The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. In The BMJ (Vol. 372). https://doi. org/10.1136/bmj. n71
Pilacuan-Bonete, L., Galindo-Villardón, P., & Delgado-Álvarez, F. (2022). HJ-Biplot as a tool to give additional analytical boost to the Latent Dirichlet Allocation (LDA) model: With an application to the analysis of digital news about COVID-19. Mathematics, 10(14), 2529. https://doi. org/10.3390/math10142529
Pons, P., & Latapy, M. (2006). Calculating communities in large networks using random walks. Journal of Graph Algorithms and Applications, 10(2). https://doi. org/10.7155/jgaa.00124
Pranckutė, R. (2021). Web of Science (WoS) and Scopus: the titans of bibliographic information in today’s academic world. In Publications (Vol. 9, Issue 1). https://doi. org/10.3390/publications9010012
R Core Team. (2020). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna Austria. URL https://www. R-project. org/.
Röder, M., Both, A., & Hinneburg, A. (2015). Exploring the space of topic coherence measures. WSDM 2015 – Proceedings of the 8th ACM International Conference on Web Search and Data Mining. https://doi. org/10.1145/2684822.2685324
Sievert, C., & Shirley, K. (2015). LDAvis: A method for visualizing and interpreting topics. https://doi. org/10.3115/v1/w14-3110
Tumushabe, J. T., Turyasingura, B., & Chavula, P. (2023). The sustainability of carbon pricing markets for climate-smart agriculture among smallholder farmers in Uganda. Asian Journal of Research in Agriculture and Forestry, 9(4). https://doi. org/10.9734/ajraf/2023/v9i4263
Vicente-Villardón, J. L. (2010). MULTBIPLOT: A package for multivariate analysis using Biplots. In Departamento de Estadística. Universidad de Salmanca.
World Bank. (2021). State and trends of carbon pricing 2021. Washington, DC: World Bank. http://hdl. handle. net/10986/41544 License: CC BY 3.0 IGO”.
Xiong, H., Cheng, Y., Zhao, W., & Liu, J. (2019). Scientific research topic analysis in the field of manufacturing using a topic model. Computers and Industrial Engineering, 135. https://doi. org/10.1016/j. cie.2019.06.010
Yue, T., Liu, H., Long, R., Chen, H., Gan, X., & Liu, J. (2020). Research trends and key points related to global carbon footprint based on bibliometric analysis: 2007–2018. In Environmental Science and Pollution Research (Vol. 27, Issue 15). https://doi. org/10.1007/s11356-020-08158-9
Zhang, L., Dong, J., Dong, Z., & Li, X. (2022). Research key points and trend analysis in the field of regional economy and carbon emissions since the 21st century: A bibliometric analysis. Sustainability (Switzerland), 14(18). https://doi. org/10.3390/su141811210.}

---

Executive summary derived from the Final Course Project of the Specialization in Data Science and Analytics of the MBA USP/Esalq

^{To learn more about the course, click here:}