![Sample our Global Development journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5929/TOC-Subject-Sample-2021-Global-Development.jpg"})
Abstract
We examine the impact of trade-related R&D spillovers from the country's partners in the MERCOSUR as well as from the European Union and NAFTA blocs and the rest of the world on total factor productivity for the Uruguayan case at the industry level, for the period 1988–95. Furthermore, we analyse the impact of domestic R&D in Uruguay. There is an evidence of trade-related technology diffusion from MERCOSUR partners to Uruguay, although domestic R&D has apositive impact on productivity. Thus, policies aimed to promote domestic R&D and decreasing trade barriers could enhance Uruguayan manufacturing productivity.
Notes
1. Knowledge diffuses across national boundaries in many ways including imports, FDI, internet, technology licensing, scientific journals and personal contacts.
2. Namely, cross-country studies do not take into account the different characteristics of countries, such as productive structure, policies and institutional settings that can affect the impact of trade liberalization and that can vary over time.
3. At the summit in La Plata, Argentina, 2005, the MERCOSUR states not to sign an FTA with the NAFTA for the time being, whereas, in Uruguay, there are ongoing discussions as to the benefits of signing a unilateral FTA with the NAFTA.
4. Trade diversion is more likely to occur if (a) the country has a tariffs on imports from the rest of the world; (b) the partner country cost is not in line with the costs and prices in the rest of the world – which would not be the case if the partner has low trade barriers with third countries.
5. The concept of dynamic efficiency gains is the same as the one in the development and productivity field, associated with a long run and cumulative improvements over time through innovation and technological modernization.
6. FDI is also related to credibility of their macro-economic policy, democracy and governance.
7. Most of the works that analyse these issues refer to the technological distance and/or absorptive capacity between domestic firms and MNEs.
8. For instance, Findlay (1978) argues that, given a certain minimum of economic development, regions or countries with a large technological gap are more likely to benefit from spillovers compared with the advanced regions.
9. In this respect, two major determinants have been emphasized: human capital (Nelson and Phelps 1966; Eaton and Kortum 1996) and domestic R&D stock (Cohen and Levinthal 1989; Griffith, Redding, and Van Reenen 2004).
10. Trade policy in Uruguay was surveyed by Vaillant (1998). See also the World Trade Organization Country Reports for Uruguay (1998).
11. See also Brazilian Institute of National Statistics and Geography (IBGE) (2011).
12. For a survey, see Schiff and Winters (2003).
13. For example, see Brown, Deardorff, and Stern (2007).
14. The rest of the world for the period analysed includes Australia, Japan, Finland, Norway and Sweden.
15. As in Olarreaga, Schiff, and Wang (2002) and Schiff and Wang (2003), R&D intensity is classified according to its share of R&D expenditures in the US. The six R&D intensive industries: 351/2: chemicals, drugs and medicines; 353/4: petroleum, refineries and products; 382: non-electrical machinery and communications; 383: electrical machinery and communication equipment; 384: transportation equipment; and 385: professional goods. The 10 low R&D intensive industries are: 31: food, beverage and tobacco; 32: textiles, apparel and leather; 33: wood products and furniture; 34: paper, paper products and printing; 355/6: rubber and plastic products; 36: non-metallic mineral products; 371: iron and steel; 372: non-ferrous metals; 381: metal products; and 39: other manufacturing industries.
16. Additionally, there are problems to merge the micro-data sets for 1988–95 and 1997–2001, because there have been important changes in the way the INE register the data, which prevent us to analyse a longer period.
17. We try pooled regression models and compared with FE by industry, and, from the comparison of the F test, we retained the FE as the more suitable ones.
18. We perform the Hausman tests to compare the FE by industry model with the FE by industry and IV methodology. Furthermore, we checked the FE IV versus the random effect IV, finding that the former was more appropriate, i.e. consistent. Nevertheless, we should note that IV estimates may not be unbiased and the goodness of fit is reduced.
19. Keller (2004) find that the contribution of own R&D explains about 50% of the total effect on productivity, whereas domestic inter-industry and foreign technology spillovers account for about 30% and 20%, respectively.
20. We also try to include a variable constructed from a principal component analysis, but this does not help us to reduce collinearity between the composite variable and domestic R&D.
21. In this regard, it is worth noting that, in 1990, 96% of the world's R&D expenditures took place in industrial countries (Keller 2002).
22. Aghion and Griffith (2005) provide a summary of recent theoretical and empirical literature on competition, entry and growth.
23. We tried as instruments domestic R&D lagged twice and once. Results are similar, so we report the results with R&D lagged once.
24. See Appendix 1.
25. The openness coefficient is defined as the total imports plus total exports of these sectors over gross output.