Patterns of technological intensity in Brazilian industry: a comparative study with developed countries.

SUMMARY

The present work aims to demonstrate that Brazilian industries display levels of efforts in technology that differ from those encountered in developed countries. In order to define these levels we have adopted indicators of technological intensity (expenditure in R&D/added value), of the structure of spending in R&D and of the human resources allocated to each industry. We have compared these indicators to those of other developed countries where similar statistics are available. The data regarding expenditure and human resources in Brazil, when compared to that of developed countries, has displayed considerable structural differences in the sectoral levels of technology intensity. These differences have been attributed to the fact that Brazil is a relatively closed market economy, with a low level of technological specialization in the high or medium-high technology industries, and except in the case of the aeronautics industry, that relies on the external flow of technology. To explain the differentiated levels that Brazil presents among its different industries, we have made use of four key factors: technology tacitness; supply's local content; the presence of foreign firms; and, the government policies adopted. Due to these differences, the present study proposes a classification of industries based on R&D intensity which, when applied to Brazil's specific case, differs from classification defined by OECD (the Organization for Economic Cooperation and Development).

KEYWORDS

industrial survey; technological intensity; sectoral patterns; international comparisons

1. INTRODUCTION

Brazil got off to a late start in its process of industrialization, which began in the 1930s. Despite the accelerated pace of growth witnessed up until 1980, the level of development in the country still falls way below the levels reached by developed countries (4 to almost 7 times less in per capita income). Industry, which directs itself essentially to attending the demands of the internal market, is made up of a steady stream of embodied and disembodied external technology flows. Even so, Brazilian industry has been making considerable efforts in technology directed towards, in most cases, adapting the flow of external knowledge to its local context. These efforts have also been brought on by local technological demands that the external flow of technology has been unable to meet. Up until now, rare have been the cases of sectors in which firms generate flows of new knowledge in order to gain dynamic competitive advantages.

The main aim of this present paper is to demonstrate that this intermediate--and relatively dependent--position bestows upon the Brazilian industries levels of technological efforts that are distinct from those witnessed in developed countries. In order to identify these levels, we have adopted indicators of R&D intensity (expenditure in R&D/ value added), of the structure of spending in R&D and of the human resources allocated to R&D, according to each sector of industry. We have compared these indicators to those of other developed countries for which there are similar statistics available. The recent study concerning technological innovation carried out by IBGE (the Brazilian Institute of Geography and Statistics), called PINTEC 2000 (Industry Technological Innovation Survey), has led to a qualitative leap in terms of data regarding technological efforts in Brazilian industrial firms. The new data are close to those encountered in other OECD countries, enabling international comparisons. This work aims to bring continuity to the articles that have sought to define levels of innovation in Brazilian industry and in other Latin American countries (Furtado et alii, 1994; Quadros et alii, 1999; Erber, 2001; Katz and Stumpo, 2001). Our intention is to put forth a form of classification of the industrial sectors according to their technological intensity that portrays the situation found in developing countries which is closer to the international dynamics of technological diffusion.

In the second section, following this introduction, a brief description will be made of the features of the methodology used in PINTEC 2000's data base. In the third item, we will analyze the data regarding technological efforts (expenditure in R&D) of the Brazilian industry, according to each sector, comparing the data to that of developed countries. After that, in the fourth part, we will build a comparative analysis between the structures regarding human resources (researchers) found in Brazil and in the United States. Then, in the fifth section, we will introduce analytical elements to explain the differences in the levels of technological efforts between developed countries and Brazil. In the sixth part we will suggest a system of classification of the sectors of Brazilian industry based on their levels of technological efforts. Finally, we will present our conclusions.

2. PINTEC 2000'S DATA BASE AND R&D INDICATORS

The Industry Technological Innovation Survey 2000 (PINTEC-2000) carried out by IBGE (The Brazilian Institute of Geography and Statistics) represents a milestone in the field of statistics regarding technological innovation in Brazil's industry. Prior to this study, an industrial technological innovation survey was carried out for the state of Sao Paulo, based on the methodology defined in the Oslo Manual, covering the period from 1994 to 1996 (Quadros et alii, 1999). This study, entitled PAEP (Survey of Economic Activity in the State of Sao Paulo'), was coordinated by the 'Seade Foundation' (The Foundation of the State System of Data Analysis) and did not bring among its results statistics regarding firms' R&D expenditures. Only the human resources allocated to this type of activity were within the study's scope.

The PINTEC-2000 was a novelty for it was the first innovation survey to be carried out on a nationwide scope, based on the Oslo Manual of OECD and on the third version of the 'Community Innovation Survey' questionnaire from the European Community. At the same time, for the very first time, systematic statistics regarding R&D expenditures of the Brazilian industrial firms were available. It was, in other words, a survey on the topic of technological innovation that brought a specific chapter about R&D, in which the firms were requested to provide data upon both internal and external R&D expenditures (see Table 1), and upon the human resources destined internally to these activities, according to their level of qualifications.

The data regarding expenditure previously available and published by the Ministry of Science and Technology was based on estimates made from incomplete data bases. This data did not allow separation of figures according to industries. Now, for the first time, a data base is available regarding Brazilian industrial firms' R&D expenditures, which allows international comparisons with OECD countries.

The scope covered by the PINTEC-2000 survey is quite outstanding and worthy of note. It represents a survey of industrial firms that have more than 10 employees, the sampling being of 11,044 firms. This sampling represents a universe of 72,005 industrial firms. Out of this total, 22,698, or rather, 31.5% claimed to have introduced technological innovations between 1998 and 2000. A subgroup of these innovators, made up of 7,413 firms (32.7%), had internal R&D expenditures in 2000. This group divides itself up into 3,178 firms that have carried out R&D activities on a regular basis while 4,235 have carried out R&D, though not in a continuous manner.

The data regarding expenditure in R&D provided by the PINTEC-2000 study is based on a universe of firms which differs slightly from the R&D figures provided in developed countries. Normally, data regarding industrial R&D is obtained through surveys, based on the Frascati Manual, which limits the survey to firms that have continuous R&D activities, i.e., at least one person working full time equivalent in these activities. The innovation surveys may eventually produce data regarding firms' R&D expenditures, but this data is not put into use in the official statistics. The two significant differences in methodologies lie in what is covered by each one. The Frascati Manual restricts the survey to firms with continuous R&D activities, whereas the surveys based upon the Oslo Manual include also firms that carry out R&D activities sporadically. On the other hand, the surveys based on the Oslo Manual only collect data regarding R&D expenditures in firms that have introduced innovations, excluding those firms that perform R&D activities without implementing innovations. On the whole, the surveys based on the Oslo Manual present a greater level of coverage than those based on the Frascati Manual but the differences in the level of expenditures are much smaller than in the number of firms (Sirilli, 1998).

3. STRUCTURAL FEATURES OF R&D EXPENDITURE IN BRAZILIAN INDUSTRY: AN INTERNATIONAL COMPARISON

By no means are the Brazilian industrial R&D activities negligible. They account for 32.7% of the domestic R&D expenditure. (1) The rest of the domestic R&D expenditure is carried out by public or private academic institutions and research centers. Although this proportion lies way below that encountered in developed countries, where it comes to exceed 75% in the United States, the efforts carried out internally by the firms in R&D activities cannot be considered irrelevant, for they define the technological level of the Brazilian industrial sector.