THE FUNDAMENTAL PATTERNS OF COMPARATIVE ADVANTAGE OF STEEL INDUSTRY IN INDONESIA Pola Dasar Keunggulan Komparatif Industri Baja Indonesia

Accurate productivity measurements can provide useful information in improving competitiveness. Therefore, it is important to understand the differences in relative productivity among countries, allowing countries to focus and specialize in their relatively more productive products. This study aims to analyze the fundamental patterns of comparative advantage, with the Indonesian steel industry as the focus of analysis. This research uses the regressionbased method of revealed comparative advantage (RCA) analysis with an instrument variable (IV) method that employs export data from 25 exporting countries to 35 destination countries during 2010 2017. The result shows that Indonesia has the strongest comparative advantage in the steel industry among the ASEAN countries. Even though the steel industry is ranked 27th in Indonesia’s comparative advantage values, several products have a strong comparative advantage and even a strong position internationally. In addition, it is worth including some ASEAN countries in the observation of estimating the key parameter of productivity, while not the main focus of the paper, yields a new estimate of θ, which is still in line with the literature.


INTRODUCTION
Pressure for protection from import competition is inevitable, and this seems especially true in the steel industry. For various reasons, the steel industry was the beneficiary of the protectionist policies in the 19th and 20th centuries. One main reason for such policies is that domestic steel production was considered crucial for state independence, as iron and steel are basic commodities and raw material for arms (Kawabata, 2018). In some countries, protection from imports was an important element of government intervention, for example, in Japan until the early 1970s and in Korea, China, and Taiwan up to the early 1990s (Lee, Ramstetter & Movshuk, 2005). Even the U.S. protected its steel industry since the 1960s (James & Parsons, 2005).
In addition to import tariffs, many instruments can be used to protect the steel industry. In the 1960s, voluntary export restraints (VERs) were used against Japanese and European steel in the U.S. steel market. The U.S. steel domestic market urged the government to take action regarding the surge in steel imports. In response, the President negotiated voluntary restraint agreements (VRAs) with Japanese and European steel producers in 1968.
These producers agreed to limit steel imports to specified maximum tonnages for a specified period (Daniel & Ross, 1989).
There was also an increase in antidumping investigations in the 1980s in the U.S. and Europe. Since then, 624 anti-dumping measures have been applied to steel products, with one-third among them imposed by the U.S.   (Demura, 1995).
Accurate productivity measurements can provide useful information in enhancing competitiveness (Tien, 2005).
By specializing in the production of relatively more productive goods, a country can gain more from trade.
Therefore, it is important to understand the differences in relative productivity among countries to allow countries to focus and specialize in their relatively more productive products.
As the Ricardian comparative advantage says, the country should produce and export relatively more in that product in which it is relatively more productive. Ricardo's main idea is that a country has a comparative advantage in a product if its relative production cost is lower than in other countries (Salvatore, 2013). In other words, the comparative advantage reflects the differences in relative productivity.  (1951,1952), Balassa (1963), andStern (1962), which showed a clear positive relationship between labor productivity and exports. They found that the industries with the higher ratios of the U.S. to U.K. exports had relatively higher productivity of labor in the U.S. than in the U.K. (Salvatore, 2013). A positive relationship between labor productivity and exports was also found in Golub and Hsieh (2000) between the U.S. and the following countries: Japan, Germany, France, U.K., Italy, Canada, Australia, Korea, and Mexico.
The most common measurement of comparative advantage is the Balassa Index of Revealed Comparative Advantage (RCA) (Balassa, 1965). The concept behind the Balassa Index of RCA is that the (unobservable) differences in relative productivity can be inferred from the (observable) pattern of trade since the pattern of trade is determined by differences in relative productivity (French, 2017). However, the Balassa Index has several empirical weaknesses, its theoretical foundation has long been debated, and its poor empirical distribution characteristics have also been criticized.
The theoretical foundation of the Balassa Index has long been debated in the literature since it does not fit the original Ricardian idea of comparative advantage (Bowen, 1983;Vollrath, 1991 In testing cross-sectional predictions, the procedure allows us to estimate the extent of intra-industry heterogeneity, typically denoted as " ." The relationship between productivity  (1) provides an unbiased estimation of , the extent of intra-industry heterogeneity in this model (Costinot et al., 2012).
Therefore, this study employs the IV method to estimate equation (1).
Technological differences are assumed to be exporter-industry specific and depend on two parameters: the fundamental productivity , , which is exporter-industry specific, and a measure of productivity dispersion , which is country invariant (Leromain & Orefice, 2013). , captures factors related to cross-country variation of productivity, such as climate, infrastructure, and institutions that affect all producers in a given country and industry.

RESULTS AND DISCUSSION
The estimation results of θ   The former difference in the symmetry of distributions can also be shown by simple density function graphs in Fig. 1. Fig. 1 illustrates the shape of the RCA distribution across 25 countries at the 2-digit level commodity over time.
The density function of the RCA index is symmetric around one (one being the threshold for having a comparative advantage in a certain sector) and very close to a normal distribution (shown in the solid line of Fig. 1).

RCA of Steel Industry
To see the patterns of the steel industry comparative advantage, the comparative advantage values for each country are presented in

RECOMMENDATION
The key structural parameter of the