This special issue of features selected articles presented a

This special issue of features selected articles presented at the 7th Sao Paulo School of Economics Conference Series, which subject was Macroeconomics in Emerging Markets. The Conference was held at the Sao Paulo School of Economics of Getulio Vargas Foundation in Sao Paulo, Brazil, on August 3, 2015. This conference was a joint effort of , which is the official journal of the Brazilian Economic Association published by Elsevier, and the Getulio Vargas Foundation. In addition to keynote speakers, the Conference featured four presentations in two special sections of articles previously selected by the Editorial Board of . These articles are scheduled to appear in this special issue.
Introduction Indeed, the Brazilian currency (real) appreciated in real terms (relative to US dollar) by 60% between 2003 and 2011. According to Armando Monteiro Neto, president of National Confederation of Industries (CNI), the appreciation of the exchange rate affects the competitiveness of the Brazilian manufacturing sector not only abroad, but also in the domestic market. Furthermore, more than 50% of all industrial segments in 2009 faced import kainate receptors (Monteiro, 2010). CANO (2012) as well as Bresser-Pereira and Marconi (2010) also emphasize this aspect, suggesting that the level of the exchange rate is the main factor behind the loss of competitiveness of the manufacturing sector during the last decade, and warn that it could lead eventually to a process of deindustrialization. There is little doubt that exchange rate has an important impact on the level of competitiveness of Brazilian manufactured goods. According to Broz and Frieden (2006) a real appreciation of exchange rate increases the purchasing power of local residents by lowering the relative price of foreign tradable goods. These authors also emphasize that there is a trade-off between competitiveness and purchasing power and that the exchange rate plays a crucial role on the determination of the equilibrium. However, it is important to point out that, in spite of its importance, the exchange rate is not the only factor that explains the loss of competitiveness of Brazilian manufactured goods. Naturally, productivity is also a key determinant of the evolution over time of the competitiveness of the domestic industry. Bonelli and Pinheiro (2012) listed direct and indirect factors related to productivity that have limited the competitiveness of the Brazilian manufacturing industry. For instance, the quality of infrastructure, the investments in R&D, absorption of foreign technology, labor cost and the educational level. As consequence, in terms of domestic factors, exchange rate is not the only key factor to explain the increase of import penetration, but productivity is also a crucial element. It is, therefore, important to examine which factor that has a larger impact in determining the share of imported goods in the total demand for manufactured goods. This article evaluates the broad effect of exchange rate and productivity on the recent dynamic of import penetration in the Brazilian demand for goods in the manufacturing sectors. Moreover, we investigate which factor has the greatest impact. At first sight, it is not clear why the impact of exchange rate and productivity on import penetration would differ. We, thus, propose a simple micro-founded model to examine under which specific conditions productivity (as measured by the ratio between industrial production and number of workers) would lead to a larger increase in import participation on domestic market when compared to an exchange rate depreciation. We show that the crucial factor that generates this difference is the participation of foreign inputs in the domestic production. This theoretical result is in line with the argument put forward by Lisboa and Pessoa (2013) that devaluations may protect national industry, but, at the same time, raise the costs of foreign inputs. To investigate whether this predicted difference in impacts is relevant empirically we make use of panel data regressions for 17 manufacturing Brazilian sector covering the period between 1996 and 2011. Since there may exist simultaneity between labor productivity, exchange rate and import penetration leading to biased estimators, we use a GMM panel data methodology based on Arellano and Bond (1991). Specifically, this method takes the first differences of the variables to eliminate unobserved sector-specific effects and use lagged instruments to correct for simultaneity. However, as pointed in several works, the use of lagged instruments in the first-difference equation may lead to a weak instruments problem, resulting in large finite-sample biases. To circumvent this problem, we also apply the system GMM panel estimation proposed by Arellano and Bover (1995) and Blundell and Bond (1998).