S&T POLICY

Print Friendly

Outliers

PINTEC data analysis indicates that cooperation between businesses and universities is only strong in certain sectors of the economy

FABRÍCIO MARQUES | ED. 254 | APRIL 2017

 

An article published in the March issue of The Journal of Technology Transfer may help explain some of the factors that affect cooperation between universities and businesses in Brazil. The study was based on responses to the 2011 Innovation Survey (PINTEC) by 17,749 Brazilian companies from a sample universe of 128,699 businesses, and showed that the number of research and development (R&D) partnerships between scientific institutions and the private sector is dependent on the intensity of innovative activities in each sector of the economy. In industries where more money is invested in innovation such as the biofuels, paper and cellulose, and electricity and gas sectors, cooperation with institutions is linked to the existence of internal R&D departments that encourage interaction with universities, as well as the availability of public financing and the size of the company (larger companies tend to collaborate more). Most of the companies that responded to the PINTEC survey operate in less technologically innovative sectors, where the logic is different: many of them do not have their own R&D departments and cooperate with universities and research institutes mainly by contracting services in response to specific needs.

“This type of partnership seems to replace the support for internal R&D activities at these companies. This finding goes against the understanding that businesses need internal R&D resources to absorb the results of external R&D,” says economist Nicholas Vonortas, one of the authors of the article. Vonortas, a professor at George Washington University (GWU) in Washington, holds a São Paulo Excellence Chair (SPEC) at the University of Campinas (UNICAMP), a FAPESP pilot program seeking to establish collaborations between institutions in the state of São Paulo and high-level researchers based abroad. The paper is co-authored by sociologist Diego Rafael de Moraes Silva, from the UNICAMP Department of Science and Technology Policy (DPCT), as part of his doctoral research, which he started in 2015. Diego is supervised by economist André Tosi Furtado, a professor at UNICAMP who also contributed to the article, and by Vonortas, from GWU. He is currently participating in a sandwich period at the GWU Institute for International Science and Technology Policy.

A primary analysis of the PINTEC data showed that a few sectors of the economy presented particularly high indicators of innovativeness in relation to staffing and R&D investment. “This was a problem, because the presence of these outliers—these sectors with standard deviations well above the average—lowers the reliability of any statistical analysis,” explains Diego. The solution was to separate the innovation-intensive outliers from the other sectors, and to study the two groups separately. This led to the conclusion that the determinants of cooperation differ in the two categories.

The data also suggests that companies more deeply involved with R&D have a much greater appetite for public funding for innovation. “This evidence indicates that rather than modifying preexisting behavior, public policy may simply be influencing sectors that would already be predisposed to innovation anyway,” says André Tosi Furtado. According to the professor, the notion that internal R&D is a valuable indicator of a company’s capacity for innovation is well established. “It is these businesses that are able to ask universities the right questions. Some sectors already operate in this way, and others appear to be moving in this direction, such as the cosmetics, electronics, optical products, and auto parts industries,” he says. Furtado observes that in countries like Brazil, where innovation systems are immature, the decision to outsource R&D is understandable. “Companies that contract R&D services from universities and research institutes often do not have the structure needed to conduct these activities internally. A large part of this cooperation is not even really R&D. It may include product testing, for example,” he says. Many business sectors in Brazil outsource R&D, and the 2014 PINTEC survey suggests that this trend may be growing in several sectors.

Points of interaction
Eduardo Motta Albuquerque, a professor at the School of Economic Sciences of the Federal University of Minas Gerais (UFMG), observes that the results of the article correspond with the analyses of other authors, who have already demonstrated that innovation-intensive companies behave in a particular way. He recently participated in an international study that highlighted the existence of 29 points of interaction between Brazilian universities and companies across 20 sectors of the Brazilian economy, while in the United States there were 47 points of interaction in 34 sectors (see Pesquisa FAPESP, issue nº 234). “But this article presents a very creative statistical and econometric analysis, highlighting a major structural problem by defining the outlier: cooperation between universities and businesses is not common in Brazil, and the few examples of such collaborations lie outside the curve—so much so that they had to be separated in the statistical analysis.”

Diego de Moraes Silva’s doctoral research has a broader ambition: to evaluate the potential of the PINTEC survey, which has been produced by the Brazilian Institute of Geography and Statistics (IBGE) since 2000, to generate new indicators and analyses capable of guiding public policy in the field of innovation. “PINTEC is great for verifying an industry’s perspective on its interactions with universities, but the abundance of data available has not been widely explored by researchers and the authorities. Public policies in general are based on R&D indicators,” he says. According to André Furtado, the questionnaire of 195 questions that companies answered in 2011 offers a vast range of data. “It is possible to conduct statistical regression analyses based on this information, which would help us better understand the phenomena of innovation in Brazil.”

Access to the PINTEC microdata is granted to researchers, but with certain restrictions. Those interested in company-level data need to submit a project to the IBGE indicating the information they need and what they plan to do with it. Because companies provide the data confidentially, the IBGE only passes it on if there is a guarantee that respondents will not be identified. Diego intends to investigate the data in detail soon, but to produce this first article, he analyzed aggregate indicators of the 55 business sectors publicly available in the 2011 PINTEC report. “The aim was to perform a preliminary assessment of whether access to aggregate data would allow us to draw interesting conclusions,” he says.

Economist and PINTEC coordinator Alessandro Pinheiro agrees that the IBGE’s results could be better explored, but he sees difficulties in expanding use of the survey, especially in academia. “Studies on the economics of innovation are generally linked to research centers that subscribe to heterodox economics and are limited to a few universities, such as UNICAMP, the Federal University of Rio de Janeiro (UFRJ), and occasionally the University of São Paulo (USP) and UFMG. Business administration and production engineering naturally focus on case studies, making little use of elaborate statistical and empirical methods,” he says.

In the context of public policy, Pinheiro states that there is a lack of knowledge in Brazil regarding the methods and tools needed to handle PINTEC’s microdata. According to him, there would be no point offering more research rooms to study PINTEC sector data at the IBGE headquarters, because the demand, mostly from state and municipal agencies, is still low. “Expertise in this field is very concentrated in São Paulo and Rio, as well as Brasília, thanks to the IPEA,” says the PINTEC manager, referring to the Institute of Applied Economic Research (IPEA), a foundation linked to the Ministry of Planning and Management whose researchers have already produced several detailed papers using PINTEC microdata.

However, Pinheiro points out that interest in the data tends to increase as the survey enters new fields, as evidenced after the addition of the gas and energy sector in 2011, or data on researcher gender and public spending in 2014. “While rarely used in scientific publications, PINTEC’s descriptive statistics are widely utilized by government ministries, science and technology departments, and private sector organizations, even providing parameters for the National Science and Technology Strategy.” According to Nicholas Vonortas, access to detailed PINTEC data promises to provide interesting results. “We are developing a plan to work with the data at the company level, and I imagine it will be highly valuable,” he says.

Project
Innovation, strategy and policy systems (nº 13/ 50524-6); Grant Mechanism São Paulo Excellence Chair (SPEC) Program; Principal Investigator Nicholas Spyridon Vonortas (George Washington University and UNICAMP); Investment R$1,655,029.25 for the entire project.

Scientific article
MORAES SILVA, D. R., FURTADO, A. T. e VONORTAS, N. S. University-industry R&D cooperation in Brazil: A sectoral approach. The Journal of Technology Transfer. Online. Mar. 3, 2017.


Matérias relacionadas

INNOVATION
Books demonstrate data about connections between universities and companies
CAREERS
Tips on how to write a good research project proposal
FUNDING
Institute to invest up to R$18 million per year in research in Brazil