Those interested in science, technology, and innovation policy (STIP) would be remiss to remove politics from the policy. With 49 percent of the world’s population heading to election polls this year, it is a real possibility that government transitions will upend countries’ current STIP trajectories. Interested stakeholders seeking metrics to track those priorities may want to read a recent article in Quantitative Science Studies, titled, “The Policy is Dead, Long Live the Policy—Revealing Science Technology and Innovation Policy Priorities and Government Transitions via Network Analysis.” Colombian research partners Julián D. Cortés and María Catalina Ramírez Cajiao analyzed how frequent and enmeshed research topics were in public funding research calls (RC) in Colombia from 2007 to 2022. Since the funding for these RCs came from public sources, they could serve as one indicator of government priorities. The researchers found that, alongside a general increase in research field diversity and density, several research fields such as drug discovery and conservation, “Maintained their higher strategic relevance despite the government in office.” If generalized, methods such as network analysis may be helpful for analysts to track science, technology, and innovation priorities across different periods of government and identify which research sectors are politicized.

Methods

Based on a literature review of STIP evaluations in Europe, Cortés and Cajiao found that the most common methods for those evaluations included, “Descriptive statistics, context, documents, and case studies.” In addition to expanding methods of analysis, Cortés and Cajiao sought to expand the research geography to lower- and middle-income countries, which they considered to be an often-highlighted but rarely addressed research gap. Cortés and Cajiao reviewed public RC data oriented toward research in Colombia’s Ministry of STI open data portal and RC digital archive. Next, they coded research fields by manually reviewing each document for what fields each RC would support. They standardized research fields by utilizing the All Science Journal Classification Codes (ASJC). For example, ASJC considers “Insect Science,” “Plant Science,” and “Soil Science,” as one overall topic, “Life Sciences.” Cortés and Cajiao did this analysis by year and matched RC priorities with periods of government (four years).

The authors also utilized co-word analysis first introduced by Callon et al. (1983) to visualize clusters of ASJC topics in trios. According to Cortés and Cajiao, “if a given RC has three ASJC, those ASJC (nodes) are collocated (linked) given that all of them are contained in the same RC.”

Figure 1: ASJC co-occurrence network (Bastian, Heymann, and Jacomy, 2009; Callon et al., 1983; DNP, 2021)

Research fields that were frequently part of ASJC co-word networks received high “Betweenness Centrality Scores,” which meant that they were research fields of interest. In contrast, research fields with lower scores may have more marginal or limited attention.

Results and Implications

From 2007 to 2022, the number of research fields in Colombia’s RCs increased. Despite changes in Colombia’s government, Physical Sciences retained its position as the top field with its high betweenness score compared the other top fields of Life, Health, and Social Sciences. Health Sciences topics are on an upward trend ever since an apparent dip in priority during the 2011–2014 government period, catching up to Life Sciences based on their betweenness scores. Despite these findings, Cortés and Cajiao caution that their research does “Not integrate the effects of STIP priority fluctuations and research/innovation outputs, nor the amount of funding by fields in the same framework.” They noted that although the Health Sciences sector’s betweenness score was not particularly impressive compared to other top fields, its research growth rate surpassed Physical, Life, and Social Sciences.

Figure 2: Number of fields Cortés and Cajiao identified in RCs with betweenness centrality score by area (left y-axis) and network density score (right y-axis) by period