S&T Policy 

Delicate balance

Studies show that the structure of BRIC scientific communities increasingly resembles the structure of scientific disciplines in developed countries

Published in August 2012

How can science and technology systems in emerging nations, such as Brazil, Russia, India and China (BRIC), be compared with those in developed countries, which have a more consolidated scientific community? A study conducted by a group of researchers from China and published on the web site of the journal Scientometrics shows that despite notable differences between the scientific structures of the so-called BRIC countries and the group of the world’s richest countries, the G7, the divide between the two blocs is decreasing. The work, whose main author is Li Ying Yang, from the National Library of the Chinese Academy of Sciences, analyzed how the scientific communities in these countries are distributed into disciplines and major fields of knowledge, using the scientific papers published by these communities as a parameter. Although scientific production from rich countries has a certain homogeneity and reflects an equilibrium among the various areas of knowledge, with life sciences leading, the BRIC countries have a more heterogeneous structure and lack a common identity, with physics, chemistry, mathematics and engineering as the dominant disciplines. The exception is Brazil, with a profile closer to that of the industrialized countries. However, this concentration has decreased over the past 20 years, with the BRIC countries becoming increasingly more balanced.

According to Yang, the decision to seek a balance between the disciplines or invest heavily in strategic areas is always complex and corresponds to the particular needs of each nation. “The disciplinary structures of each country are influenced by cultural factors, political history and their geography. They are also affected by scientific and technological development,” the researcher observed. “Experience shows, however, that an asymmetric structure can harm sustained development in science and technology,” wrote Yang, referring to the importance of maintaining a scientific community interested in all fields of knowledge and prepared to face future scientific and technological challenges wherever they appear.

The Chinese group developed a methodology for comparing the structure of disciplines in science in the BRIC bloc and the G7. The papers published on Thomson Reuters’ Web of Science database in 1991, 2000 and 2009 furnished the raw material for the analysis. These papers were classified according to the list of more than 170 disciplines in the Journal Citation Reports, which is used to assess scientific journals. It should be made clear that the methodology uses just the number of papers and not the number of citations and/or impact factors. Its concern, therefore, is quantity, not quality. Likewise, it ignores the weight of international collaborations. Each paper is credited to just one country, the country of the corresponding author (the person responsible for submitting the work for publication).

Two parameters were used to develop an index for the analysis. First, the percentage of the number of papers from a country in each discipline relative to the total number of papers from the same country was determined. In addition, the percentage of papers in each discipline in the world relative to the total number of papers published worldwide was calculated. The ratio between these two parameters was used to define a standardized indicator, the Activity Index (AI), which measured the weight of each discipline in the country relative to the world and helped the researchers compare the disciplinary structures of the countries.

The paper contains interesting graphs, such as those that illustrate the opening of this report. In one of the graphs, the disciplines were regrouped into four major areas of knowledge: life sciences; agrarian sciences; earth, environment and energy sciences; and mathematics, physics, chemistry and engineering. The 10 most prominent disciplines in each country, those that had the highest Activity Index, were evaluated. Although its disciplinary structure is considered to be balanced, the United States had just a single color graph – all 10 disciplines with the greatest weight were from the life sciences, such as medicine and biology. In China and India, however, the 10 disciplines belonged to the fields of mathematics, physics, chemistry and engineering. Brazil had seven disciplines in life sciences, one in mathematics, physics, chemistry and engineering, and two in agrarian sciences.

Finally, the Gini coefficient for each country was calculated from the distribution of the weight of the disciplines, as measured by the AI. The Gini coefficient is used to measure inequality. If it is lower than 0.3, the authors consider that the discipline structure is balanced. If it exceeds 0.4, they consider that the structure is asymmetric and polarized. The evolution of the Gini coefficient between 1991 and 2009 was able to show clearly how each country is changing. The BRIC countries are in the process of reducing concentration and seeking equilibrium. Brazil, for example, had a coefficient of 0.568 in 1991, an indication of a polarized structure, but this index decreased to 0.389 in 2009, an indication of equilibrium. A similar trend was found for India, a decrease from 0.471 to 0.360, and for China, a decrease from 0.568 to 0.395. Russia was the great exception. The Gini coefficient has remained at approximately 0.64 since 1991. Investment in disciplines in the fields of mathematics, physics, chemistry and engineering, which dates to the Cold War era, is still prevalent. The United States increased its Gini coefficient slightly, from 0.182 in 1991 to 0.229 in 2009. In France, inequality decreased – the coefficient changed from 0.297 to 0.158. “The developed world is standing still, while we’re changing,” says Hernan Chaimovich, vice president of the Brazilian Academy of Sciences, a retired professor of the Chemistry Institute of the University of São Paulo and an adviser to FAPESP’s Scientific Department.

If the equilibrium trend among these areas of knowledge is encouraging, the data motivate a discussion of the desirability of particular varieties of fine tuning among the disciplines, although a model that proves successful for one country is not necessarily suitable for another nation. “In the United States, there’s fierce debate about the disciplines’ relative loss of space over the last few years in the field of engineering, including a drop in the number of undergraduate students, but it’s impossible to say that this discussion applies to other countries,” says Rogério Meneghini, scientific coordinator of the SciELO electronic library. In the case of Brazil, a lack of engineers is already being identified as an obstacle to sustained economic growth (see Pesquisa FAPESP 149). According to Meneghini, unlike the other BRIC nations, Brazil has a tradition in life science disciplines. “This goes back to the start of the 20th century, with great scientists such as Carlos Chagas and Oswaldo Cruz, who were heavily influenced by research that was being done in Europe. But China, India and Russia did not come under this influence,” he says. He mentions the example of genetics studies in the former Soviet Union, which were stifled for a substantial part of the past century due to the influence in the country of the Ukrainian biologist Trofim Lysenko (1898-1976), who would have nothing to do with Mendelian genetics.

For Hernan Chaimovich, the debate about this fine-tuning in Brazil is relevant but still secondary. “The number of scientists in the country is still small when compared with the average number in developed countries. We need to increase the number of scientists in all areas,” he says. The fundamental question, according to Chaimovich, is that the quality of Brazilian research is not increasing in the same proportion as scientific production. “Quantity and quality need to grow together. We have to create strategies that favor this,” says the professor, who cites as an example FAPESP’s strategy for encouraging the internationalization of Brazilian science by funding research projects in collaboration with institutions from other countries.

Although developed countries have a more consolidated structure, it is a mistake to imagine that they form a monolith. Another study published online in the same journal, Scientometrics, in March and authored by Peter Schulz and Edmilson Manganote, professors from the Gleb Wataghin Physics Institute at the State University of Campinas (UNICAMP), shows that the so-called “Western model,” in which there is a predominance of medical and biomedical research, has in fact a series of nuances. The United States and United Kingdom, in fact, have a very similar pattern, with medicine representing one-quarter of their scientific production, but the distribution is different in countries in continental Europe, with a slightly greater share in physics and chemistry and with medicine varying between 18% (Spain) and 30% (Austria). “The Western model is clearly divided into two sub-groups,” says Peter Schulz.

The researchers from UNICAMP based their analysis on the profiles of countries, as published by Science Watch, from Thomson Reuters, which ranked the accumulated scientific production of each country (including citations) for approximately 10 years in the 22 main fields of knowledge. Based on these data, the Brazilians proposed a new indicator, the country profile index (CPI), also with the objective of comparing the disciplinary structure of countries. “The article by the Chinese uses a more sophisticated methodology. In our case, we initially used open data, because we were thinking of creating an analysis tool aimed at an audience that is not very specialized in scientometrics, but that would be an important agent in decision-making in the scientific territory,” explains Schulz. The CPI also shows the heterogeneity of the profile of the BRIC countries and indicates that the profile of publications in Brazil is similar to that of Argentina and Mexico, with a predominance of clinical medicine, chemistry, physics, botany and zoology, and with the agrarian sciences occupying a greater space than the world average (4% of the total Brazilian production). South Korea, Taiwan and Japan form another clearly defined bloc, with a prominent role for engineering (see graphs). “Disciplinary structure can be correlated with the economic development strategies in each country, as we briefly indicated in the case of the Asian tigers and the BRIC countries,” says Schulz.

Both the Chinese and the Brazilian studies suggest a change in the profile of disciplines in Brazil, bringing it ever closer to the Western model. Manganote from UNICAMP warns about a bias in these data, the recent inclusion of several Brazilian scientific publications on the Web of Science database. There were just 26 Brazilian journals on this database in 2006. Today, there are more than 100. “The spectrum of indexed Brazilian publications on the WoS database expanded after 2007. But this group certainly doesn’t yet correspond to the Brazilian reality. Some areas, like the agrarian sciences, business and economics, are probably still concealed by the incompleteness of the basis used,” he says, referring to Brazilian production that is still not indexed internationally. Another important bias, say the researchers from UNICAMP, concerns the low weight of human and social sciences in the disciplinary structure in various countries, including European countries. This is a result of the reality that scientific production in these areas is not represented in journals indexed on Web of Science but in regional publications and in books. “In our paper, we discuss how the WoS database doesn’t provide an entirely true picture of the disciplinary structure of a country,” says Peter Schulz.