Keywords

Introduction

International trade has novel issues of new protectionism. In line with the decrease in tariffs for imported goods, some World Trade Organization (WTO) member states introduce a new policy that seems to become a protectionist policy for their domestic industry. The protection against “price discrimination” and “pricing strategy” activities—that which importers carry out to gain market power are also known as dumping (Gifford and Kudrle 2009)—is an instance of such policy. After being ratified in 1994 by WTO, the anti-dumping policy has been implemented by many of the member states. In the Anti-dumping Agreement (1994), a practice of international trade could be categorized as dumping if the selling price of the product sold to the country of destination is lower than the market price in the country of origin. If the country of destination (the importer) is able to prove that its domestic industry was harmed by the dumping practices, the importer can exercise anti-dumping measures against the goods sold by the exporter. Even so, researchers have different opinion about the anti-dumping policy.

The anti-dumping policy is a form of new protectionism that is widely used to protect domestic producers from “unfair” international trade practice. Bown suggests that the anti-dumping policy is generally issued by developed countries such as the United States, the European Union (EU), Canada, and Australia. However, since the beginning of 2000, the anti-dumping policy was also initiated by developing countries, including Indonesia.

Experts are divided over the effectiveness of anti-dumping policy. Viner (1923) and Marsh (1998) argue that anti-dumping policy is an efficient instrument to protect domestic companies from pricing strategy undertaken by the company abroad. The addition of anti-dumping tariffs is considered able to restore prices of exported goods from the country impacted by anti-dumping duties (ADs) (named country) to the reasonable market price so that domestic products can be competitive in domestic market. At the interim, Prusa (1999) is still not convinced of the effectiveness of the anti-dumping policy considering there are unintentional trade effects that may occur due to the implementation of the anti-dumping policy.

One of the unintentional trade effects that may result from the implementation of the anti-dumping policy is the presence of trade diversion (Bown and Crowley 2007). Domestic consumers who previously traded more efficiently with the named countries, but due to the imposition of anti-dumping tariffs, have lost supplies from the named countries and are forced to switch to other trade partner that is less efficient. For instance, after the United Kingdom imposed anti-dumping duties on New Zealand butter, it switched to Danish butter which was priced higher.

Protection measures that importing countries can opt are actually not only in the form of anti-dumping duties. There are such other options as countervailing, quota restrictions, and safeguards. Nevertheless, anti-dumping is considered a better protectionist policy as Tharakan (1995) argues. He suggests that anti-dumping duties are better than quota restrictions since anti-dumping duties will not increase additional cost when demand increases, while in quota restrictions the extra cost will be even greater due to the increased consumers’ demand. On the other hand, safeguards and countervailing are deemed expensive policies that pose difficulties in the implementation. Anti-dumping is considered easier to do so; it is more often chosen to anticipate “unfair” international trade (Neufeld and CNUCED 2001; Tharakan 1995).

However, anti-dumping is believed to be able of giving effect in the future when the countries accused of dumping take countermeasures. Moreover, anti-dumping tends to drive price increase in domestic market, thus reducing domestic consumers’ welfare. As Zarnic and Vandenbussche point out, trade barrier causes domino effect, that is exports re-route by exporting countries, which could eventually lead to implementation of similar protection by the new export destination countries. This situation is hurting the exporting country.

Developing countries have different characteristics with developed countries that have become the subject of research on the impact of anti-dumping. According to www.worldsteel.org data, Indonesia in the year 2015 was in the 14th rank among national steel importers in the world by value of imports which amounted to 11.4 million tons, while the European Union and the United States rank the 1st and 2nd with the value of imports reaching 37.7 million tons and 36.5 million tons, respectively. Indonesia has very low average steel consumption per capita compared to that of developed countries such as the United States and the European Union. Although the trend has increased, Indonesia’s consumption in 2007 only touched 37.3 kg per capita before jumping to 58.3 kg per capita in 2016. The figure is far below the United States which in 2007 consumed 400.5 kg per capita and the European Union which took up 448.1 kg per capita until 2016. The consumption trend decreased until the year 2016 when the US consumption stood at 318.4 kg per capita and EU consumption hit 338.7 kg per capita.

With such different characters, anti-dumping policy may potentially have different impacts for Indonesia’s context. The authors therefore attempt to undertake research on anti-dumping in such developing countries as Indonesia. As for the types of goods that dominate the dumping investigation in 2014 were metal, especially steel. Indonesia is chosen as its government has imposed anti-dumping measures on four types of carbon steel products since the period 2010–2013, namely H&I Section, Hot Rolled Coil (HRC), Hot Rolled Plate (HRP), and Cold Rolled Sheet/Coil (CRC).

Literature Review

Conceptualization

Indonesia’s steel industry is currently categorized as a strategic manufacturing industry by the Ministry of Industry. Nevertheless, it is one of the industries with weak competitiveness according to the Ministry’s criteria. This encourages the government to facilitate the industry to be able to compete with imported products. Along with the increasing number of demands in iron and steel products that are not accompanied by production capacity and production cost efficiency, it is feared that the industry would not be able to compete with imported products to meet domestic needs. This is evident in the deficit of net export of iron and steel products which amounts to 9.5 million tons (USD9.4 million) in the year 2014.

On the other hand, China becomes the world’s largest iron and steel manufacturer that is able to generate a total production of 823 million tons (2014) or approximately 49% of the total world production (source: World Steel Organization) and posts a sizeable value of iron and steel net exports accounting for 78 million tons (2014) or USD 33 million (2014). With the slowdown in China’s economy, iron and steel products are currently experiencing supply (crude) excess. This excess is further coupled with government policies which no longer give incentives to the construction sector so as to force domestic manufacturers to export its products. According to the World Steel Organization data, in 2015 China exported 112 million tons, with 20–50% cheaper price than that of any other producer (source: Forbes Asia).

Facing such unfavorable situation, the Indonesian steel industry filed an anti-dumping petition to Indonesian Anti-Dumping Committee (KADI) against several steel products from such countries as China, Taiwan, Korea, Japan, Malaysia, Singapore, Vietnam, and Ukraine in the period of 2008 until 2011. The petition was approved and enacted as Minister of Finance Decree following a lengthy (which could take up to two years) investigation. Nonetheless, despite the anti-dumping ruling, Indonesian steel producers still complain about the flooding imports of steel. This would become an interesting phenomenon if trade diversion occurs amid the anti-dumping policy, which will then be deemed ineffective. The situation would be more alarming as the United States recently confirmed anti-dumping duties on Chinese steel products. A trade diversion from the US anti-dumping measures, if eventually takes place, would make Indonesia’s steel market further flooded by imported products.

It is worth noting that the anti-dumping policy in Indonesia mostly impacts steel products if viewed from the imports value side. As many as 78.3% of imports value from products that are subject to AD consist of steel products which vary from HRC, CRC, HRP, to H&I Section. This research will focus on anti-dumping policy in those four products within the period 2008–2015. It is questionable whether the anti-dumping policies are successfully tapping steel imports or causing trade diversion as demonstrated in the preceding studies countries conducted in such developed nations as those of the European Union and the United States.

Trade diversion that may occur as the impact of the anti-dumping policy takes form as the migration of the source of imported goods from the countries that are subject to anti-dumping tariffs to other countries that are not. In numerous studies, the exploring countries that are hit by AD is referred to as “named countries”, while other countries that are not impacted by AD are often referred to as “non-named countries”. Table 8.1 illustrates that the named countries for HRC products are Korea and Malaysia, while the remaining 43 countries (not hit by AD) are non-named countries. Prior researches managed to identify the presence of trade diversion from named countries to non-named countries due to the anti-dumping policy imposed by the developed countries. The authors intend to find out whether trade diversion also takes place in developing countries such as Indonesia.

Table 8.1 Indonesia’s anti-dumping actions against imports of steel products
Full size table

Previous Empirical Research

Neufeld and CNUCED (2001) argues that anti-dumping is a good instrument to reduce the negative impact in a free market regime. If this opinion is true, with anti-dumping measures, the competition between the domestic industry and the exporting countries would become neutral again. In reality, however, an anti-dumping action in one country is continuously followed by similar measure in other countries. In a developing country such as India, Ganguli (2008) finds, although still in early phase of investigation, a volume drop of steel imports by up to 50%. It seems that the announcement of dumping investigation already shocks the importers so that they respond by lowering the imports volume. Similar results are also found in the research by Vandenbussche and Zanardi (2010).

As discussed earlier, the study seeks to identify one of the unintentional trade effects from the imposition of anti-dumping policy, that is trade diversion. There are sizeable researches in trade diversion as the result of anti-dumping policy. Subsequent researches on the impact of anti-dumping are conducted by adding more details of categorization of named countries and non-named countries. Applying this model, Prusa (1996) finds that AD in the United States manages to limit the volume of imports from the named countries, but causes shift of imports to non-named countries. This shift is defined as “trade diversion” in his research. Even though trade diversion occurs so that the total volume of imports does not change much, at least an anti-dumping policy is still able to raise the selling price of imported goods so as to make the domestic industry able to compete.

Other than that of Prusa (1999), another study that identifies “trade diversion” from named countries to non-named countries is conducted by Konings et al. (2001). This research takes European Union as object and finds out that the trade diversion in European Union is not as strong as that found in the United States as discussed in Prusa’s study. In this research, the size of the diversion seems to be influenced by the amount of the anti-dumping tariffs. The results of this research are supported by Brenton (2001), who also examines the trade diversion in European Union and identifies a trade diversion from named countries to non-named countries.

Inspired by the results of research by Konings et al. (2001), Mendieta performs similar research in Mexico utilizing the same method. His research reveals that the volume of imports from named countries fall and a diversion of volume to non-named countries like Indonesia takes form. Since the volume of imports has not changed much post imposition of anti-dumping, Mendieta opines that anti-dumping policy is not effective in controlling the volume of imports.

Moreover, Bown and Crowley (2007) perform research to unmask whether a trade diversion occurs from the anti-dumping policy on Japan, while Park Soon-Chan (2009) does similar study for the case of China. Employing ordinary least squares (OLS) and generalized method of moments (GMM) methods, both studies demonstrate that in addition to trade diversion stemming from anti-dumping actions on China and Japan, a trade depression—which is the decrease in exports value of named countries as a whole in the global market—also come off.

The trade depression phenomenon is also captured in the research by Egger and Nelson (2011) and Prusa (2013). Egger and Nelson use panel data and gravity model data for the period of 1948 up to 2001. The results suggest that anti-dumping measures give a negative effect on the international trade in the world because it significantly lowers the total value of world trade, thereby reducing the people’s welfare. Besides revealing trade depression, Prusa (2013) also discovers the fact that the impact of imports value drop is greater at the time of the initial dumping investigation than it is at the time of the anti-dumping policy decision.

Researches on the impact of anti-dumping policy on trade are quite plentiful and broad, yet there has been no consensus on what kind of impact that emerges. Some of the above researches (Prusa 1999; Brenton 2001; Konings et al. 2001) discover that anti-dumping policy is effective in either limiting or reducing imports from the named countries. Nonetheless, it remains dubious whether the imports shift to non-named countries or affect domestic production, provided that the empirical results depend upon which country that imposes anti-dumping policy and which industry that is hit by such policy.

An industry-based research is performed by Lee and Jun (2004) utilizing data of anti-dumping policy in the United States to observe whether trade diversion occurs in the chemical, metal casting, and steel industries. The study stumbles on the presence of trade diversion in the chemical and metal casting industries, but not in the steel industry. Another similar study is also carried out by Durling and Prusa (2006) that applies world trade data of HRC steel products for the period 1996–2001. Trade destruction stemming from the anti-dumping policy is identified but not trade diversion. However, when the data are limited to only the United States, empirical results indicate trade diversion.

From the earlier elaboration, it can be implied that this study attempts to contribute to the literature that discusses the impact of the anti-dumping policy on the importation of HRC, CRC, HRP, and H&I section. The impact of the anti-dumping policy which is observed includes:

  1. 1.

    Trade reduction/restriction of named countries due to AD on steel products in Indonesia

  2. 2.

    Trade diversion from non-named countries due to AD on steel products in Indonesia

Hypothesis

Pursuant to theories and prior empirical researches, the hypothesis proposed in this research is the occurrence of trade reduction/restriction from named countries and the existence of trade diversion to non-named countries because of AD on steel products in Indonesia.

Research Methods

Data Source

The data that will be used in this quantitative research are taken from transactional Notice of Imports (PIB) sourced from the Directorate General (DG) of Customs and Excise, Ministry of Finance. This research will look into the data from the year 2007, which is the year in which dumping investigation on HRC products from Korea and Malaysia started, until the year 2015 or two years after imposition of AD on CRC products from China, Taiwan, Korea, Japan, and Vietnam. The GDP data are obtained from the World Bank.

Based on the available data, this research employs sample of transaction data of steel imports in Indonesia for the period 2007–2015 considering that the anti-dumping policy was issued in that period. Nevertheless, since the amount of data between individuals is not identical, the unbalanced panel is used. The sample countries are listed in Table 8.2.

Table 8.2 Sample countries
Full size table

Empirical Specification

The econometrics model that is utilized is multiple linear regression OLS, referring to the model specifications by Konings et al. (2001) combined with variable control used by Bown and Crowley (2007). In the model by Konings et al. (2001), a variable Number (dummy number of named countries with Number = 1 if the number of named countries for certain HS products at certain times is three or more) is used, but the authors apply variable market share (dummy percentage of the total number of imports from named countries compared to total imports with market share = 1 if percentage of imports from named countries reaches 50% or more). This is because the use of reference of three countries is less consistent. For instance, although there are three countries that impose AD for HRP, those countries only constitute 32% of Indonesian HRP imports.

The variables of exporting country’s GDP and productivity of the domestic industry as discussed in the research by Bown and Crowley (2007) are used as independent variables. The GDP variable is utilized as their research takes the United States and Japan as objects that own production network between industries of the two countries. As the researchers do not see clearly the existence of production network between Indonesia and steel exporting countries, the researchers do not use the exporter’s GDP variable. On the other hand, the productivity of the steel industry does not become the focus of this research, so that the domestic steel industry’s productivity is not used as an independent variable.

$$ {\displaystyle \begin{array}{l}\ln {M}_{jit}={\alpha}_0+{\beta}_1\ln {M}_{ijt-1}+{\beta}_2{\mathrm{AD}}_{ijt}+{\beta}_3{\mathrm{Named}}_{it}+{\beta}_4{\mathrm{Marketshare}}_{ijt}\\ {}\kern5.5em +{\beta}_5\ln \kern0.28em {Y}_t+{\beta}_6\ln \kern0.28em {e}_{ijt}+{\beta}_7\ln \kern0.28em {p}_{ijt}+{\varepsilon}_{ijt}\end{array}} $$

In the above equation, i represents HS steel product code, j represents country (named or non-named country), t represents period, α0 represents constant intercept from parameter estimate,β1 to β9 represent slope of parameter estimate, while εijt represent error term. For more details, Table 8.3 is a description of the variables used in this study.

Table 8.3 Variable and data source
Full size table

Methods of Estimation

This research applies OLS with panel data in the estimation. Yet, due to the different number of observations in each country, it is called unbalanced panel. Greene states that by using panel data, there is a flexibility to model behavior differences between individuals. In addition, by having more unit cross section, data panel can minimize biased results. Due to the use of dummy named/non-named country in this study, the study is approached with fixed-effects model. The fixed-effects model utilizes constants as parameters of the regression (intercept). Assuming there are no statistical problems, the fixed effect can be estimated consistently so that estimation parameters depend on the impact of country/region and year in samples (Hsiao 2004).

Results and Analysis

Description of Statistics

Description of statistics based on the observed data is depicted in Table 8.4 as follows:

Table 8.4 Statistical description
Full size table

Table 8.4 clearly indicates that the smallest imports volume per transaction recorded in the Directorate General of Customs and Excise is only 4.97e−06 m3, that is, transaction of imports of CRC 7209160010 HS steel type from Japan, dated 26 December 2007. Meanwhile, the largest transaction is 72,789.73 m3, that is, transaction of imports of HS 7209170010 CRC from Vietnam on 20 March 2013. This is considered large since the transaction is subject to AD.

There are 18 currencies used by the 48 countries which export steel to Indonesia in the period 2007–2015. Interestingly, given the minimum value of exchange rate is Rp1, there are import transactions that use rupiah, particularly 2308 transactions throughout 2007–2015. The countries that trade in rupiah are Japan, Singapore, China, and Malaysia.

Estimation Results

Based on estimation results of the fixed-effects model, Table 8.5 indicates significant negative AD tariff. This is in accordance with some prior researches carried out in developed countries which suggest that the imposition of AD is able to degrade the value of Indonesia’s imports from named countries. Meanwhile, the results of the estimation show that the variable of named country is positive. In contrast to the studies by Konings et al. (2001) and Prusa (1999), this research offers empirical evidence of absence of trade diversion amid anti-dumping policy on steel products in Indonesia.

Table 8.5 Estimation results of fixed-effects model
Full size table

The unproven occurrence of trade diversion might be attributable to the named countries which become the majority exporters of CRC steel (above 50%). Indonesia relies on the supply of CRC steel from China, Taiwan, Korea, Japan, and Vietnam. When AD hits those countries, importers will still purchase the products from those countries, but certainly reduce the volume, considering the higher price. Despite the AD tariffs, the price of steel products from the named countries average only Rp2.1 billion per m3, while the average price of steel products from non-named countries reaches Rp2.8 billion rupiah per m3. Most transactions (more than 76%) are still valued below the average price of steel products from non-named countries. This could be the reason for Indonesia to keep taking supplies from the named countries.

Failing to prove the existence of trade diversion, the research identifies the presence of trade depression. Trade depression is one of the unintentional effects from anti-dumping policy as Bown and Crowley (2007) described: imports volume drop is not only experienced by the named countries, but also by the non-named countries. Figure 4.1 depicts that after May 2013 (anti-dumping on CRC was imposed on 19 March 2013), the imports value of steel in Indonesia from both the named and non-named countries experience a declining trend. The negative dummy market share variable signifies that the imports volume from the named countries that constitute the majority of steel exporters to Indonesia is lower when compared to the imports volume from the named countries that are not majority exporters.

The Indonesia’s GDP has significantly positive effect on the value of imports as it serves as a proxy of the country’s buying power. Since rising purchasing power can be coupled with the growing demand for products from other countries, it is reasonable if the rise of GDP is followed by the increase in steel imports volume.

The currency exchange rate is significantly negative. The negative sign on the exchange rate variable means that the weaker the rupiah, the lower the imports value. The weakened exchange rate reduces the purchasing power or makes transaction value higher than before so that demand for steel imports is also down.

The last control variable in this estimation is the price of the product. Significant negative variable of price means that the higher the price, the lower the volume of imports. Such price change will lower the purchasing power and eventually reduce the capability of importing steels.

Conclusion

The research unveils many interesting findings, one of which is the use of rupiah currency in import transactions, while typically in international trade, the seller’s currency or widely used currencies are used. What makes this finding more interesting is rupiah currency is used by such countries as Japan and China which dominate import transactions and make up a huge market share in the total steel imports. Therefore, this research suggests that if the use of rupiah in import transactions could be expanded to other import transactions, Indonesia’s economy would reap many benefits, one of which is adding the country’s international reserves.

As in the preceding researches, the Indonesia’s anti-dumping policy on steel products from the eight named countries are proven to cause reduction in the trade volume from those countries. This finding reinforces the series of empirical results on trade reduction/destruction as the impact of anti-dumping policy. These results indicate that the anti-dumping policy is an effective trade barrier.

The presence or absence of trade diversion as the impact of anti-dumping policy remains interesting topic. Some researches manage to prove it, while others do not. The results of this study support the opinion that the presence of trade diversion stemming from anti-dumping policy cannot be proven. There is no empirical evidence of imports volume increase from non-named countries due to the imposition of AD on imported steel products in Indonesia.

Moreover, other data show a declining trend of volume of imports from non-named countries. This could mean that the anti-dumping policy could cause trade depression in Indonesia. The anti-dumping measures on steel not only succeed in lowering the volume of imports of steel from named countries, but also cause the volume of steel imports from non-named countries to fall. The declining trend of imports volume in total post-enactment of AD policy on 19 March 2013 goes against the growing trend of demand for steel that the country is experiencing. Further research is required to find out whether the domestic industry is able to fill the demand gap or there are issues of circumvention, that is, turning HS carbon steel into HS boron steel, as has been lately suspected by the US steel industry.

It is necessary to evaluate the implementation of government’s AD policies by considering unintentional effects such as trade diversion and trade depression as well as circumvention before continuing or extending the duration of AD policies. Although in this study trade diversion is not proven, the government still needs to look back the AD tariffs, particularly since this research reveals that the selling prices of steel from the named countries are lower than those of the same products from non-named countries.