Over the last two decades, a new reform agenda towards School Autonomy with Accountability (SAWA) has spread globally, transforming school governance around the world (Verger et al., 2019). The SAWA agenda aims to transfer decision-making from central levels to schools while establishing accountability mechanisms and common standards as monitoring instruments of school performance (Verger et al., 2019). Advocated by OECD (2011) and the World Bank (Arcia et al., 2011), both developed and developing countries have progressively adopted the main tenets of this reform. In this context, Colombia arises as one of the few Latin American countries that has transformed its system along the SAWA agenda. However, what differentiates Colombia's case from others is its piecemeal and incremental approach: the reform was progressively adopted over the last 20 years and throughout three different waves marked by three presidential administrations: under Álvaro Uribe (2002-2010), Juan Manuel Santos (2010-2018) and Iván Duque (2018-2022). 

The study of policy diffusion has gained large scholarly attention across different disciplines, as globalization has accelerated the spread of global reforms (Wimmer, 2021). What puzzles scholars researching traveling reforms is why countries from different regions, with divergent institutional trajectories or inscribed in varying contexts seem to adopt similar policies (Steiner-Khamsi, 2004). Policy diffusion has been explained by different and sometimes overlapping mechanisms, such as competition among countries, coercion from international organizations, normative emulation of global scripts, or policy learning from ‘best practices’ (Dobbin et al., 2007). In particular international organizations have been central to diffusion studies in education, considered carriers of global templates (Ramirez, 2012), or been responsible for transformations due to aid conditionality (Hossain, 2022). 

At the same time, policymaking has moved towards evidence-based regulation, which means showing that decisions are not purely politically driven but also evidence-based (Maroy, 2012). A growing body of literature has looked into the knowledge architecture behind policy reforms (Baek et al., 2020), as well as the role of international organizations as knowledge brokers for policy diffusion (Waldow & Steiner-Khamsi, 2019). Arguably, the choice of certain evidence and knowledge sources is indicative of important dimensions of the policy process, such as the problem-framing and selection of potential solutions (Haas, 1992), neglected by scholars who have focused on policy coalitions (Kingdon, 1984), actors' motivations, and interests (Howlett & Ramesh, 2003) or supranational coercion or emulation (Dobbin et al., 2007) when studying reforms adoption. Clearly, what sources are chosen from all available data and what actually counts as evidence and knowledge can uncover the ideological affiliation, sources of legitimacy, and policy preferences behind a policy reform adoption.  

In this particular case, the first question I explore is what type of knowledge has Colombia used through the years to justify the adoption of SAWA? In other words, what has been, if any, the linkage between the different administrations when planning and implementing changes in education? Secondly, in the context of a global movement towards evidence-based policy, how has the use of knowledge changed during the last 20 years? To answer these questions, I aim to explore the knowledge sources used by Colombia’s policymakers over the years to justify and inform the reform adoption as well as how these sources have shaped the Colombian policy discourse around education. By studying the use of knowledge and the authorship of those sources, I intend to contribute to the understanding of a key dimension behind the policy diffusion process like the role of ‘reference societies’ (Steiner-Khamsi, 2016) and ‘sources of legitimacy’ (Edwards et al., 2018), as well as their role in the context of global policy diffusion.  often steered as key determinants of reform adoption.

Colombia's policymaking is a highly technical and hierarchical procedure, where different government agencies create policy and research reports, long-term planning documents, and policy programs. Hence, to answer the research questions, I looked into all the education-related documents from Colombia’s government for the 2002-2022 period and analyzed their respective citations. In this case, similar to Baek et al., (2018), I focused on ‘official policy knowledge’, using documents published by the National Council of Social and Economic Policy, the National Planning Department, and the National Ministry of Education. In total, I retrieved 25 documents. From each of the documents, I coded all the references and authors into a single database. I entered a total of 1233 citations divided into three reform waves linked to the different administrations: 552 citations from wave I, 177 from wave II, and 504 from wave III. In addition to the quantitative analysis of measuring the frequency of citations, I coded different attributes for all documents: (i) year of publication, (ii) publisher or institutional affiliation of the author, (iii) location of publication, and if authors or publishers were international government organizations (IGOs). From the database, I have created a text-based network analysis (Borgatti et al., 2013) to examine both the social structure of policy discourse and to interpret the different resulting knowledge networks. To analyze the data, I used the software program UCINET 6.289 (Borgatti, Everett, & Freeman, 2002) to create the database and generate descriptive statistics. Then, the program NetDraw 2.097 enabled me to visualize the relationships between the documents in the data set. I created a 2-modes network of documents and their references, followed by a 2-mode network of documents and authors. The rationale for going beyond documents and creating an authors-documents network lies in the key role actors have in policy discourse formation. In the context of Colombia’s incremental adoption of the SAWA agenda advanced by the OECD and the World Bank and the fact that Colombia became an OECD member in 2018, one would expect a growing presence of these IGOs throughout the documents. Yet, the interest is not just the frequency with which these or other authors are cited, but also how important they are in the context of the knowledge architecture of Colombia’s ecosystem. For this, I calculated an ‘in-degree’ centrality measure from both authors and documents. This measure captures the total of incoming citations for a given author or document.

In line with the movement towards evidence-based regulation (Maroy, 2012), Colombia’s policy documents relied more frequently on citations across the reform waves. Where in the first wave (2002-2010) each document had on average 45 citations, whereas on the third wave (2018-2022), the citations average was 126. In addition, all documents in the last reform wave had a separate reference list section on top of footnotes, which didn’t happen in the two first periods. The network of the source-documents and their respective cited documents shows an incohesive network with clusters of sources and cited documents in each reform wave. First, this means that each reform period draws knowledge from highly specialized sources. Second, only a few citations bridge the different reform waves. In spite of the lack of connexions across reform periods, Colombia’s overall direction moved towards the incremental adoption and consolidation of the SAWA agenda throughout different administrations. When compared to the network of source documents and cited authors, this network is not only more dense and cohesive but also shows a high number of authors being repeatedly cited across different source documents and reform waves. First, Colombia’s government bodies rank at the top of cited authors, showing a clear focus on its own knowledge to justify and create reforms. Second, the OECD ranks fourth as the most cited author, appearing in all reform periods and more often since 2014, after Colombia’s accession process started in 2013. Lastly, degree centrality shows the National Planning Department and the National Ministry of Education as the most central actors, followed by the OECD and the World Bank. These initial findings highlight the importance of both domestic and global sources in policy diffusion while further content analysis of most cited documents will reveal new insights about the knowledge used for SAWA adoption in Colombia.

Arcia, G., Macdonald, K., Patrinos, H. A., & Porta, E. (2011). School Autonomy and Accountability. World Bank. https://openknowledge.worldbank.org/handle/10986/21546 Baek, C., Hörmann, B., Karseth, B., Pizmony-Levy, O., Sivesind, K., & Steiner-Khamsi, G. (2018). Policy learning in Norwegian school reform: A social network analysis of the 2020 incremental reform. Nordic Journal of Studies in Educational Policy, 4(1), 24–37. Borgatti, S. P., Everett, M. G., & Johnson, J. C. (2013). Analyzings Social Networks. Routledge. Dobbin, F., Simmons, B., & Garrett, G. (2007). The global diffusion of public policies: Social construction, coercion, competition, or learning? Annual Review of Sociology, 33, 449–472. Edwards, D. B., Okitsu, T., Da Costa, R., & Kitamura, Y. (2018). Organizational legitimacy in the global education policy field: Learning from UNESCO and the global monitoring report. Comparative Education Review, 62(1), 31–63. https://doi.org/10.1086/695440 Haas, P. M. (1992). Introduction: Epistemic Communities and International Policy Coordination. International Organization, 46(1), 1–35. Hossain, M. (2022). Diffusing ‘“ Destandardization ”’ Reforms across Educational Systems in Low- and Middle- Income Countries: The Case of the World Bank , 1965 to 2020. https://doi.org/10.1177/00380407221109209 Howlett, M., & Ramesh, M. (2003). Agenda-Setting: Policy determinants, policy ideas, and policy windows. In M. Howlett & M. Ramesh, Studying Public Policy. Policy Cycles and Policy Subsystems (pp. 120–142). Oxford University Press. Kingdon, J. W. (1984). Agendas, alternatives, and public policies. Little, Brown. OECD. (2011). School Autonomy and Accountability: Are They Related to Student Performance? OECD. Ramirez, F. O. (2012). The world society perspective: Concepts, assumptions, and strategies. Comparative Education, 48(4), 423–439. https://doi.org/10.1080/03050068.2012.693374 Steiner-Khamsi, G. (Ed.). (2004). The global politics of educational borrowing and lending. Teachers College, Columbia University. Steiner-Khamsi, G. (2016). Comparing the Receptions and Translations of Global Education Policy, Understanding the Logic of Educational Systems. In T. D. Jules (Ed.), The Global Educational Policy Environment in the Fouth Industrial Revolution (Vol. 26, pp. 35–57). Emerald Group Publishing Limited. Verger, A., Fontdevila, C., & Parceriza, L. (2019). Constructing School Autonomy with Accountability as a Global Policy Model: A Focus on OECD’s Governance Mechanisms. In The OECD’s Historical Rise in Education. Waldow, F., & Steiner-Khamsi, G. (2019). Understanding PISA’s Attractiveness: Critical Analyses in Comparative Policy Studies. Bloomsbury Publishing Plc. Wimmer, A. (2021). Domains of Diffusion: How Culture and Institutions Travel around the World and with What Consequences. American Journal of Sociology, 126(6), 1389–1438.