Literature review

As highlighted by Goldberg and Pavcnik (2004), the discourse of the general equilibrium

in the final goods sectors. Feenstra and Hanson (1996, 2003) later tested this assumption and pointed out the important role played by the trade in intermediate goods in increasing the demands for skilled workers in both developed and developing countries. They argued that the increase of global production networks or outsourcing activities do play an important role in increasing the demand for skilled workers in both developed and developing economies.

According to Goldberg and Pavcnik (2004) the main argument presented in Feenstra and Hanson is based on the assumption that manufacturing of final goods requires the input of intermediate goods which are different in terms of their skill intensities. Trade liberalization and removal of capital controls transfer the production activities of intermediate products from the developed countries to developing countries. On the one hand these intermediate products are categorized as goods produced by highly unskilled labors from a developed country`s point of view. On the other hand they are seen as being produced by highly skilled labors from a developing country`s view point. Due to these assumptions, there has been an increase in skill intensity in both types of countries, which lead to an increase in skill premiums for both developing and developed economies.

Another argument based on the framework of explanation above is the impact of Foreign Direct Investments (FDIs) in increasing the demand for skilled workers in developing countries.

This argument is not only focusing on the production of intermediate goods. Cragg and Epelbaum (1996), and Behrman, Birdsall, and Szekely (2000) argued that if skilled labors are complements of capital goods, cheaper prices of these capital goods which resulted from trade liberalization, will lead to an increased demand for both the capital goods and the skilled labors which are seen as complementary goods for the other.

Another channel which can result in Skill-Biased Technological Change or accelerate SBTC was recently suggested by Acemoglu in 2003. Using his model of endogenous technological change, he debated that for developing countries a different definition of technological change should be allotted. For these countries this change could mean importing more machineries and equipments from other advanced countries to complement the skilled workers to improve production processes.

Acemoglu further explained that trade reforms influences the numbers of skilled labors hired through the decreasing of prices of capital goods. This in turn will increase their imports.

Empirically speaking this framework has two important features: a) imports of new machineries and equipments from developed economies should be increased following the trade reform process of that particular developing nation; b) for those sectors importing machinery and equipments from other developed nations they should see a dramatic increase in the hiring of skilled labors in those sectors. Another mechanism was also introduced by Aghion, Burgess, Redding and Zilibotti (2003) on how trade reforms can influence technological or technical change. The measure of technological change is then used to measure wage inequality. They hypothesized this by using their plants` model. They suggested that plants that are falling behind in technological advancements are more likely to be unable to compete with foreign competitors, while plants that are up to date with the technological advancements in the sector may be more likely to succeed in warding off foreign competitors penetrating the market. The authors also highlighted how institutions, labor market laws and the rate of adopting new technologies can impact the trade policies in a trade liberalized country. For the case of Brazil, Araujo, Bogliacino, and Vivarelli (2011) have undergone a study using panel data obtained from the manufacturing sub-sector in the country from years 1997-2005. They also concluded that SET effects are evident for Brazil and emphasized that technology has cause the skill-upgrading shift among workers in the manufacturing sub-sector.

In 1999, Harrison and Hanson published their analysis on the trade reforms process that took place in Mexico in the 80s. They discovered that for plants in the manufacturing sub-sector, the plants importing new machinery, equipment and intermediate products are also prone to employing more skilled labors if compared to other plants not importing any of those products mentioned. In the case of Chile it is quite different as highlighted by Pavcnik (2003) who suggested that in the beginning of the 80s, the higher numbers of skilled-labors employed in their plants is not cause by the adoption of imported intermediary goods of other technical assistance given by their foreign business counterparts.

The empirical analysis of the relationship between SBTC effects and inequality are quite

for improvements. In a related paper Aghion, Burgess, Redding and Zilibotti (2003) used a 3-digit level manufacturing sector dataset to analyze the case of India`s trade reforms episode to prove this hypothesis. They found that profit margins and the level of productivity were increasing for the plants that were up to date with the productivity enhancing methods in their sector. They also found that those plants located in states with more unrestricted labor laws observed the same gains. The differences of the degree of impact introduced by the trade reform policies, in the manufacturing sub-sector presents the wage gap between the skilled and the unskilled workers. The nearness of a plant to different types of labor restrictions and access to new technologies are important in determining the degree of wage inequality between the skilled and unskilled workers in the manufacturing sub-sector.

According to Vivarelli (2013) based on traditional view in economics, unemployment caused by technology is a temporary situation. This then can be automatically compensated by the mechanisms of the market which aims to reintegrate the employees who had lost their jobs.

These mechanisms are called the compensation theory, which was presented by Marx in his discussions on large-scale industry and the introduction of machinery (see Marx 1961: Chap. 15).

According (Vivarelli, 1995; Pianta, 2005) there are six compensation mechanisms which work to offset technology's labor-saving effects. These are; (1) hiring additional workers in the capital goods sector where new machineries and equipments are being manufactured, (2) lowering prices as a result of lower production costs due to technological innovations, (3) making new investments by using extra profits incurred as a result of technological change, (4) lowering wages which results from price adjustment mechanisms and this leads to higher levels of employment, (5) increasing income which is the result of redistribution of gains from technological innovation, and (6) producing new products by using new source of technologies.

Unfortunately, the analysis to quantify the effectiveness of the compensation mechanisms mentioned in the previous paragraph is not a direct one. This is the same for measuring the quantitative impact of technology on employment as a whole. These two points have been the central issue of a controversial debate among labor economists according to (Vivarelli, 2013) who has been publishing papers on this topic since 1995. He continued to explain that low demand of labor and capital/labor substitution elasticity, unemployment, pessimistic expectations

and delays in making investment decisions that are included in that compensation could only partially explain the impact. The discussion and debate on compensation mechanisms and how they have been functioning have often occurred when we observe the developed nations. For the case of developing countries, the validity of this debate is even more questionable because these countries are seen as passive recipients of technologies from other developed nations. Besides looking at the quantitative impact of technology on employments of skilled and unskilled workers, there have been various studies which have qualitatively analyzed the relationship of technology and employment. The analysis has produced a new concept of Skill Biased Technological Change (SBTC).

It was Griliches (1969) and Welch (1970) who first developed this concept. It was created based on the hypothesis of capital skill complementarity, which assumes that employers will increase the demand of skilled workers due to the usage of new technologies. The use of these technologies, are being adopted especially in modernized industries where only skilled workers can operate the machinery and equipments (Machin, 2003). Due to the lack of related theories explaining the SBTC phenomenon, the studies on SBTC are therefore of an empirical nature.

The phenomenon was first observed when in the past 30 years SBTC has gained momentum resulting from the increase use of computers and the influence of information technology in specific sectors (Pianta, 2005). Berman, Bound and Griliches (1994) was the first to study the impact of SBTC empirically. They highlighted that there was a strong correlation between within industry skill upgrading and increased investment in both computer technology and R&D in the U.S. manufacturing sector between 1979 and 1989. Later focusing in the US, Autor, Katz and Krueger (1998) also observed the wide spread use of computer technology since 1970. They highlighted that SBTC explains from about 30 to 50 percent of increases in the growth rate of relative demand of skilled labor in the US. Besides those researches Machin and Van Reenen (1998) found strong evidence in support of the SBTC assumption when they undergo a cross-country research for seven countries in the OECD bloc. They found that there is a positive relationship between expenditure of R&D and relative demand for skilled workers.