Introduction
Research context and problem statement
In recent decades, trade has experienced remarkable growth, as evidenced by statistics from the World Trade Organization (WTO) highlighting a significant upward trend in the export and import of merchandise products This surge underscores the increasing importance of international economics, with many economists identifying trade as a key driver of economic growth (Frankel & Romer, 1999; Irwin & Terviử, 2002; Wacziarg & Welch).
Trade openness enhances economic efficiency through labor division, allowing firms to benefit from economies of scale and modern technology via foreign partnerships It attracts capital investment to low-capital countries, driven by technological advancements that reduce transportation and communication costs Additionally, governments are increasingly adopting policies that promote integration, such as bilateral and multilateral trade agreements, to boost international trade However, despite the overall trend toward openness, some nations implement temporary barriers like antidumping measures to protect domestic producers in the short term.
Table 1.1 Merchandise exports by selected economy (USA, Canada, Mexico, Brazil, Argentina, China, Japan, India, Australia and New Zealand, South Africa, Germany, United Kingdom, France, Italia) in billion USD
Source: author collected from WTO report
Antidumping duties (AD) were relatively rare in the 1960s, averaging only 10 cases per year (Schott, 1994) However, their prevalence surged in the 1970s and 1980s as countries favored AD over tariffs, due to the more flexible interpretation of AD rules under GATT/WTO (Blonigen & Prusa, 2001) The 1980s alone saw over 1,600 AD petitions filed (Finger & Artis, 1993), and the use of AD by WTO members tripled from the early 1980s to the late 1990s (Prusa, 2005) Following the 2008 global crisis, developed countries increasingly turned to AD as a means to protect domestic producers (Kee, Neagu, & Nicita, 2010) The motivations for implementing AD include the clear GATT/WTO rules that facilitate tariff elimination or reduction, as well as the significant rise in bilateral and multilateral trade agreements that impose constraints on tariff increases, leading to heightened tensions among domestic suppliers Additionally, there is a notable lack of alternative solutions for protection aside from AD, given the conditions set by GATT/WTO and preferential trade agreements.
AD are flexible and easy to interpret (Hansen & Prusa, 1995).
Numerous studies examine the impact of antidumping (AD) measures on target countries Prusa
This paper examines how enterprises in developing countries respond to antidumping (AD) measures from developed nations, with a focus on the European Union's duties on Vietnamese footwear Since its economic reforms in 1986, Vietnam has experienced a significant increase in trade with the global market However, the country has faced numerous AD investigations, with over 40 cases reported by 2011, resulting in taxes imposed on a third of these cases Notably, the US initiated AD measures on Vietnamese catfish in 2002 and shrimp in 2003, while the EU imposed taxes on bicycles in 2004 The EU remains one of Vietnam's key trading partners, alongside the US and China.
Between 1995 and 2004, the European Union (EU) saw a significant increase in trade, with textiles and footwear being two of Vietnam's primary exports In July 2005, the EU initiated an antidumping duty on Vietnamese and Chinese footwear, specifically targeting products classified under code 6403 (CN8), which includes those with rubber or plastic outer soles and uppers The tax rate started at 4.2% in April 2006 and escalated to 16.8% by September of the same year, leading to a notable decline in footwear exports to the EU This research examines how firms are adjusting their strategies to mitigate the adverse effects of rising taxes on their exports.
Table 1.2 Vietnam Merchandise Trade in percent over GDP
Source: author collected from World Bank Data Indicators
Table 1.3 Vietnam export to EU in million USD
Source: author collected from Vietnam General Statistics Office
Vietnam's footwear industry serves as a prime example of how firms in developing countries adapt to anti-dumping duties The European Union represents a significant market, accounting for 65% of Vietnam's total footwear exports, highlighting the industry's importance to the nation's economy.
The significant sector driving Vietnam's exports and job creation faces challenges in quantifying the effects of protectionism, such as tariffs or anti-dumping measures, due to difficulties in interpreting ex-post data (P Krugman et al., 1995) However, the availability of firm-level data collected by the General Statistics Office (GSO) during 2003, 2004, 2006, and 2007 provides an ideal case for analysis, allowing for a comprehensive examination of the impacts before and after the imposition of taxes on both control groups and affected firms.
“treatment” group In addition to firm-level data, we also have sufficient products level data at 8-digit CN codes and 10 digit HTS codes level from Eurostat and USITC, respectively.
On those given, it would be concluded that unique and ideal opportunity to identify the firm’s adjustments under antidumping case.
Research objectives
Explore how the Vietnam footwear industry and Vietnam footwear firms response to antidumping duty from EU.
Research questions
1 How much does EU AD alleviate to export value, quantity and price of Vietnam footwear industry?
2 How much does EU AD reduce revenue of Vietnam footwear firms?
3 How much does AD effect on the firm’s input decision: firm’s size and labor payroll?
4 How protectionist measure in EU can adversely affect another economy (US)?
Research methodology, data and scope
Double differences (DD) method is applied to estimate effect of AD on Vietnam footwear DD is frequently used for evaluate impact of a policy or program It compares
This study examines the "control" and "treatment" groups before and after the intervention of policy, leveraging the advantages of eliminating time-invariant factors To address potential selection biases, we incorporate firm characteristics and select a control group from light manufacturing industries, such as textiles, paper, and wood, to ensure the parallel assumption of the Difference-in-Differences (DD) approach is met The data for this research is sourced from the Vietnam Enterprise Surveys, collected by the Vietnam General Statistics Office (GSO) and its provincial sub-institutions Additionally, we utilize product-level data at the 8-digit CN codes and 10-digit HTS codes from Eurostat and USITC With the investigation of Anti-Dumping (AD) measures starting in 2005 and tax increases occurring in 2006, our analysis employs panel data from both the ex-ante and ex-post periods, as detailed in Chapter 3.
Thesis structure
This paper is structured into five chapters Chapter I introduces the objectives and research questions, highlighting the significance of the study Chapter II reviews existing research, covering both theoretical and empirical backgrounds Chapter III presents an overview of the impact of EU anti-dumping measures on the Vietnamese footwear industry, along with a summary of relevant data Chapter IV details the research methodology and its findings, while Chapter V concludes with a brief summary of the paper's key points.
Literature Review
Theoretical Review
Free trade generally benefits economies, but it creates both winners and losers, prompting some countries to implement protective policies for their industries These policies often aim to achieve not just economic objectives but also political ones Key trade policies adopted in recent decades include import tariffs, export subsidies, and dumping practices A significant measure within the World Trade Organization framework is the antidumping duty, which provides essential protection for local producers in a timely manner.
2.1.1 Market analysis in context of Trade in a single industry
This paper utilizes the theory of partial equilibrium, referencing Feenstra (2003) and Krugman & Obstfeld (2008), to analyze two markets: Home and Foreign In this scenario, both markets consist of producers and consumers for a single product, A, with prices quoted in Home currency It is assumed that the price of product A in the Home country (P) is higher than the price in the Foreign market (P*), resulting in Home importing product A from Foreign.
To have transparent view, we use supply and demand curve as figure 1.
The domestic market's demand and supply curves are illustrated in Figure 1a, while Figure 1b presents the Home import demand curve (MD), derived from the previous figure Home import demand is determined by the Home demand adjusted for Home supply At the price P0, the demand is represented as D0, with domestic supply at S0, resulting in a specific quantity demanded for imported goods.
When the price rises from P0 to P1, the demand for homes decreases to D1, while the supply increases to S1 This shift results in a decline in home import demand to D1 - S1 Generally, the import demand curve slopes downward, indicating that as prices increase, the quantity demanded decreases.
Import demand is equal to zero at the equilibrium price of Domestic supply and demand, PE,
At this price, Home country does not need to import goods.
The foreign export supply curve, illustrated in Figure 2, shows the relationship between price and export quantity In Figure 2a, at price P0, foreign suppliers produce S0 while demand is at D0, necessitating an export of D0f - S0f As the price rises to P1, foreign consumption decreases to D1 and supply increases to S1, leading to a higher export quantity of D1f - S1f This indicates that the export supply curve (XS) has a positive slope, as higher prices incentivize foreign producers to export more However, at price PE, there is no motivation to export, as all goods are consumed domestically, resulting in zero export quantity.
Figure 2.2 Foreign Export Supply Curve
2.1.2.1 Effects of tariff on Volume and Price
The home government imposes a specific tax (t) on each unit of product A imported Without this tariff, the price of product A in the domestic market aligns with the world price (PE) However, with the introduction of the tariff, mediators are only incentivized to import products from abroad if the price difference between the domestic and foreign markets is at least equal to the tariff amount This creates an excess supply in the foreign market and a shortage in the domestic market, leading to price adjustments Consequently, the price of product A in the home market rises while the price in the foreign market falls, ultimately resulting in a price difference that equals the tariff (t).
Home Market World Market Foreign Market
Quantity, Q Qt Qe Quantity, Q Quantity, Q
Figure 2.3 Impact of tariff on import price and volume
The introduction of a tariff t leads to an increase in Home prices to Pt, which is higher than the original price P, prompting Home producers to increase production Consequently, the demand for imports decreases, moving from point E to point A as shown in figure 2.3 In Foreign, the price drops from P to Pf = Pt – t, resulting in a reduction of exports from point E to point B on the XS curve and an increase in demand due to the lower price Overall, the traded quantity decreases from Qe to Qt, establishing a new equilibrium where Home's import demand and Foreign's export supply meet at Qt, with Home prices remaining higher than Foreign prices.
In economic terms, a country is classified as "small" if its demand for a product does not affect the world price, while a "large" country is one whose demand does have an impact on global pricing This distinction is important for understanding the effects of tariffs on social welfare, as the implications can vary significantly based on the size classification of the country involved.
2.1.2.2Consumer surplus and producer surplus
A tariff impacts both domestic and foreign prices, leading to an increase in domestic prices while foreign prices decrease As a result, domestic consumers face losses, whereas domestic producers benefit from the tariff Conversely, foreign consumers gain from lower prices, but foreign producers experience losses To effectively compare the costs and benefits of tariffs on these distinct groups, it is essential to quantify them using the concepts of consumer and producer surplus, which are fundamental in microeconomics.
Consumer surplus refers to the benefit consumers receive when they pay less for a product than the maximum price they are willing to pay For example, if a customer is willing to spend $100 on a pair of shoes but buys them for $80, the consumer surplus is $20 On the other hand, producer surplus is the profit that producers earn when they sell a product for a higher price than the minimum price they are willing to accept.
A product might be sold with 15$ while manufacturer is willing to sell with 12$, then 3$ is producer surplus Total consumer surplus and producer surplus is called as social welfare.
2.1.2.3Effects of tariff on Social welfare
The impact of tariffs on the social welfare of a large importing country can be significant When a tariff is imposed, domestic prices rise from P to Pt, while foreign prices decrease to Pf This leads to an increase in domestic supply from S1 to S2 and a decline in demand from D1 to D2 The producer surplus, which is the area above the supply curve and below the price curve, increases by area a due to the price increase Conversely, consumer surplus, defined as the area above the price and below the demand curve, decreases by the sum of areas (a + b + c + d) The government benefits from tariff revenue, calculated as (D2-S2)*t = c + e Assuming this revenue is used to fund public services, the net social welfare can be expressed as producer benefits plus government tariff revenue minus consumer losses, resulting in e - (b + d) The rectangle e represents the terms of trade gains from being a large market with market power, while the triangles b and d indicate efficiency losses caused by market distortion due to the tariff.
Figure 2.4 Effect of tariff on social welfare in case of large country
In the case of a small importing country, as illustrated in Figure 2.5, the country's tariff does not influence the world price, leading to a domestic price of P0 + t, where P0 represents the price prior to taxation This scenario results in zero terms of trade gains and an efficiency loss represented by areas b and d Therefore, the net welfare of the small importing country is calculated as - (b + d).
Figure 2.5 Effect of tariff on Home demand in case of small country
Research by Brander and Spencer (1992, 1984) and Dixit (1984) indicates that government-imposed tariffs, whether specific or ad valorem, lead to increased home prices Their findings suggest that in large home countries, a portion of the tax burden may be passed on to domestic consumers while benefiting foreign producers Additionally, Krugman (1979) provides evidence supporting these concepts within a monopolistic framework These studies form a robust theoretical foundation for this paper.
Export subsidies are financial incentives provided by governments to promote exports, which can take the form of specific or ad valorem subsidies These subsidies increase domestic prices for exporters while lowering prices for importers, creating a gap equal to the subsidy amount As a result, consumers incur losses, while producers benefit, and the government bears the cost of the subsidy The overall net welfare impact is negative due to efficiency losses caused by distorted behaviors among producers and consumers Additionally, if the exporting country influences global prices, export subsidies may lead to a decline in world prices, resulting in further terms of trade losses, although this effect is negligible for small countries.
Figure 2.6 Effect of export subsidy on export country
The Ricardian labor model emphasizes labor productivity and comparative advantages, as discussed by Akerlof & Yellen (1986) and Feenstra (2003) In this model, we represent labor productivity in the Home country as ai for industry i, while ai* denotes the equivalent productivity in Foreign countries The wages in the Home and Foreign countries are indicated by w and w*, respectively Consequently, the unit labor cost for industry i in the Domestic market is expressed as w/ai Home country holds a comparative advantage in industry i when the condition w/ai < w*/ai* is satisfied, indicating lower unit labor costs relative to Foreign.
Empirical Review
Numerous empirical studies have examined the topic of antidumping, with a comprehensive summary provided by Blonigen and Prusa (2001) Pierce (2011) indicates that antidumping measures can reduce physical productivity in the host country, prolonging the existence of low-productivity firms and ultimately decreasing resource efficiency Additionally, Egger and Nelson (2011) utilize panel data from 1948 to 2001 to assess the effects of antidumping on trade volume and welfare within the GATT/WTO framework, revealing that while the negative impact exists, it is relatively modest Prusa (2006) highlights that the East Asia region experiences a higher number of antidumping duties compared to other regions Furthermore, Anderson, Schmitt, and Thisse (1995) demonstrate through the Bertrand-Cournot model that antidumping policies diminish consumer surplus while enhancing domestic producer surplus, leading to a conflict where consumers tend to oppose antidumping laws while producers advocate for higher duties Ultimately, the government aims to maximize total domestic surplus, suggesting that antidumping laws may inadvertently increase global surplus.
Exploring the determinants of Anti-Dumping (AD) measures is a compelling area of research Numerous studies have investigated the motivations behind AD actions A key factor is the clear guidelines set by GATT/WTO aimed at eliminating or reducing tariffs, which subsequently boosts imports and intensifies competition for domestic suppliers Additionally, there is a notable absence of effective alternatives to safeguard domestic industries against the impacts of increased imports.
In addition, GATT/WTO conditions to raise AD are flexible and easy to interpret (Hansen & Prusa, 1995), As discussed by (Blonigen & Prusa, 2001; Herander & Schwartz, 1984; Sabry,
Three primary factors influence antidumping (AD) filings: employment levels in the affected sector, the capital intensity of that sector, and the impact of foreign product competition Blonigen and Prusa (2001) conducted a comprehensive review of earlier research, while the other two studies utilized industry-level data and single equations to derive their conclusions.
Studies on the impact of antidumping (AD) measures highlight significant effects on foreign firms and exports Prusa (1994) found that AD initiates pressure on foreign companies to raise prices during the initial stages of tax petitions Cuyvers and Dumont (2005) demonstrated that Asian exports to the EU decreased substantially in both quantity and value following the filing of petitions Notably, from 1981 to 2001, 74% of European Commission petition investigations resulted in either increased taxes or price undertakings Brambilla et al (2012) examined the AD imposed by the USA on Vietnamese catfish, revealing an 85% reduction in imports two years post-tax, leading to income declines for catfish households ranging from 36.7% to 74%, depending on their initial income ratio Furthermore, investment in catfish farming plummeted between 28.3% and 61.9% Prusa (2003) also confirmed these adverse outcomes.
AD causes a 30-50% declines of import In general, previous studies are quite consistent in belief that AD might harm import.
Anti-dumping (AD) measures create discrimination between named and non-named countries, often leading to trade diversion towards non-named firms, particularly evident in the EU (Brenton, 2001) Research indicates a significant decline in imports from named countries to India, while imports from non-named countries have increased (Ganguli, 2008) Similarly, studies on China reveal evidence of trade depression and diversion attributed to rising prices of “dumped” products (Park, 2009) In the USA, strong evidence of trade diversion has also been documented (Krupp & Pollard, 1996; Prusa, 1996) Conversely, Konings et al (2001) found limited trade diversion effects in the EU from 1985-1990, although this conclusion was challenged by later research using extensive panel data (Konings & Vandenbussche, 2005) If trade diversion is minimal, AD measures can significantly enhance domestic firm markups Despite numerous studies on import diversion, there is a lack of research on the export diversification of named countries, which is critical since AD is based on product and market discrimination Consequently, foreign countries may diversify their products or markets to circumvent these trade barriers.
Economists generally agree that anti-dumping (AD) measures negatively impact foreign countries, particularly developing nations, by raising the prices of imported products and reducing demand The benefits of AD for domestic firms are often unclear and depend on trade diversion effects Most existing research relies on aggregate or industry-level data, leaving a gap in firm-level analysis of AD's impact on foreign enterprises This paper addresses this gap by utilizing unique ex-ante and ex-post Enterprise Survey data from Vietnam to assess the effects of European Commission AD measures on Vietnam's footwear industry Additionally, there is a scarcity of studies examining the market diversion effects of AD, and our analysis also explores how protectionist measures in the European Union can adversely affect the United States economy.
Conceptual Framework
Given theories and empirical studies indicate a conceptual framework, which is demonstrated in figure 1.9, for this study.
Market 2 increases Antidumping Market 1 declines
Price Export to Market 1 increases
Effect on Vietnamese Footwear Firms
Market 2 increases Value Export to Market 1 decreases
Total export to market 1 and market
Antidumping duties lead to a reduction in exports from foreign countries within targeted industries This phenomenon is supported by both theoretical and practical evidence Theoretical frameworks, as discussed in section II.1.2, highlight insights from Akerlof and Yellen (1986) and Feenstra (2003), which effectively explain the impact of antidumping duties on the exports of specific industries.
In addition to volume, economists also agree with impact of AD on import price (Akerlof &
Yellen, 1986; J Brander & Spencer, 1992; J A Brander, 1981; A.J Brander & Spencer,
Research consistently distinguishes between two scenarios regarding import countries: (1) small countries and (2) large countries In the case of small import countries, anti-dumping (AD) measures may have little to no impact on foreign prices Conversely, for large import countries, AD actions can lead to a significant decrease in import prices This distinction is supported by numerous empirical studies, including works by Brambilla et al (2012), Cuyvers & Dumont (2005), and Prusa (1994, 1996, 2003).
The decline in import volumes may prompt foreign firms to take strategic actions, such as diversifying their markets or products to circumvent tariffs While research specifically addressing this phenomenon is limited, numerous studies have explored trade diversion towards non-targeted countries (Brenton, 2001; Ganguli, 2008; Park, 2009) This anti-dumping (AD) measure can lead to a significant increase in the market share of non-designated countries within the home market, suggesting a potential shift in trade dynamics.
Antidumping (AD) measures can lead targeted firms to shift their focus towards new markets or similar products that are not affected by tariffs If firms are unable to identify alternative markets, they may experience a decline in both output and revenue According to Mankiw (2012), there is a strong correlation between output and input, meaning that a reduction in output often forces firms to decrease their production inputs, such as workforce size and labor wages, in order to optimize profits.
Textile, sewing products Rice Footwear Fishery productsWood and wooden products
An overview of Vietnamese Footwear Industry
EU tax rising on Vietnamese footwear
Vietnam's footwear industry is a significant contributor to the country's exports, with a reported export value of $2.5 billion in 2004, ranking third after textiles and rice The Vietnam Ministry of Industry and Trade indicates that approximately 500,000 workers are directly employed in this sector, primarily from agricultural backgrounds, many of whom lack college degrees and earn low wages Footwear enterprises are concentrated in major cities such as Hanoi, Ho Chi Minh City, and Da Nang Vietnam produces a diverse range of footwear products, including leather and sports shoes, which are known for their quality and affordability due to low labor costs As a result, these products have gained substantial acceptance in markets across the EU, US, and Japan.
Figure 3.1 Top five Vietnamese export goods exclude oil in 2004
Source: author collected data from Vietnam General Statistics Office
China ,People's Republic ofVietnam India Indonesia Tunisia
Figure 3.2.Top five footwear exporters to EU
Source: author collected data from EU Trade market access database
The European Union (EU) is the largest importer of Vietnamese footwear, with imports valued at approximately €2,198,294,000 in 2004, equivalent to nearly $1.77 million USD, which represents 68% of Vietnam's total footwear export value According to EU Commission trade data, Vietnam ranks as the second-largest footwear exporter after China, yet it faces significant competition from countries like China, India, Indonesia, and Tunisia Additionally, the Vietnamese footwear industry primarily relies on imported materials and machinery A key product in this sector is footwear classified under code 6403 in the 8-digit Combined Nomenclature (CN8), featuring outer soles or uppers made of leather.
On July 7, 2005, the European Commission announced the initiation of an anti-dumping investigation into leather shoes imported from Vietnam and China, following a complaint from the European Confederation of the Footwear Industry (CEC), which represents 40% of EU suppliers The investigation is based on Article 5 of European Regulation 384/96, as amended by Regulation EC No 384/96 (461/2004), focusing on footwear categorized under CN8 code 64032000.
The investigation, covering the period from April 1, 2004, to March 31, 2005, scrutinized a total of 15 firms, including eight from Vietnam and twelve from China, with 163 and 86 respondents respectively The findings were published in (EC) No 553/2006, revealing the tax implications outlined in Table 3.1.
Table 3.1 EU AD on Vietnam and People Republic China Footwear
The tax requisition significantly impacts the export activities of both countries to the EU As illustrated in Figure 3.3, the volume of products under EU investigation has notably declined since 2005 In 2003, Vietnam's AD footwear exports to the EU experienced a remarkable increase of 15% in value and 35.5% in quantity, a trend that continued in 2004 with growth rates of 9.2% in value and 15.8% in quantity However, this upward trajectory has since changed.
The European Union's announcement of anti-dumping (AD) measures led to a significant decline in value, dropping to 0% in 2005, -8.5% in 2006, and -13.8% in 2007, accompanied by quantity changes of -5.5%, -12.78%, and -13.05% respectively These trends align with the theoretical findings of Dale (1980), Feenstra (2003), and Staiger & Wolak (1994), as well as empirical evidence from Brenton (2001) and Messerlin (1989) Furthermore, Prusa (2003) supports these observations, noting a 30% to 50% reduction in AD product imports to the US, corroborated by Brenton's (2001) report of a 20% decrease in the first year following the implementation of similar tax measures.
Figure 3.3 EU import Vietnam AD footwear and non-AD footwear
Source: author collected data from http://epp.eurostat.ec.europa.eu/newxtweb/
Data Description
My analysis bases on three dataset: EU import data, Vietnam Enterprise Survey and
US import data This section gives a description for each dataset.
3.2.1 EU Import 8-digit (CN8) Data
Data on EU footwear imports is sourced from the European Commission's EUROSTAT website, which provides comprehensive export and import statistics for EU member states, defined as EU25 This includes countries such as Belgium, Denmark, France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain, the United Kingdom, Austria, Finland, Sweden, Cyprus, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Slovakia, and Slovenia The analysis utilizes 8-digit Combined Nomenclature (CN8) data, specifically for antidumping product classification Import data is presented in terms of value (in Euros) and quantity (in kilograms), excluding tariffs, freight, insurance, and other surcharges The price is calculated by dividing the value by the volume, with data spanning from 1999 to 2012.
Table 3.2.Vietnam AD footwear export value, volume, price to US compareto control groups
Groups Obs Mean Std.dev Min Max Growth
Quantity in 100 kgValue in thousand Euro
Table 3.2 presents a comparative analysis of the import value, volume, and price of Vietnamese AD footwear to the EU, alongside similar data from Thailand and Indonesia, as well as Vietnamese apparel exports to the EU The data is averaged over two distinct periods: prior to 2004 and post-2006, following the announcement of tax changes in July 2005 and the subsequent lifting of AD at the end of that year.
Between 1999 and 2011, Vietnam's AD footwear export value increased by only 8.2%, reaching 28,842 thousand euros, while Thailand and Indonesia experienced significant growth rates of 47.6% and 77.8%, respectively India, used by the EU as a benchmark for assessing Vietnam, saw an impressive increase of 146.6% In contrast, the average growth rate for 8-digit apparel, a comparable product, was 91.5% Vietnam's AD footwear quantity declined by 6.2%, while Thailand, Indonesia, and India recorded substantial growth rates of 44.7%, 21.8%, and 104.8%, respectively Notably, Vietnam's AD footwear prices rose by 24.1%, the highest among these countries Overall, the data indicates a significant decline in Vietnam's AD footwear volume and value exports to the EU compared to other countries and the apparel sector.
Table 3.3 provides insights into Vietnam's Non-AD footwear products compared to those from India, Indonesia, and Thailand The EU has announced a petition and imposed taxes on 31 product lines, while 75 footwear lines at the 8-digit level are involved, with 44 lines exempt from penalties Following the duty imposition, Vietnam's Non-AD footwear exports increased by 6.6%, contrasting with significant declines in Thailand (60.8%), Indonesia (21.8%), and India (0.6%) However, Vietnam experienced a drop in both quantity (15.5%) and export price (37.2%) During the same period, Thailand and Indonesia also saw decreases in export quantity by 61.3% and 35.3%, respectively, while India recorded a 2.6% increase In terms of pricing, India and Indonesia experienced slight increases of 9.9% and 6.5%, respectively, while Thailand saw a minor reduction of 5.3%.
Table 3.3.Vietnam Non- AD footwear export value, volume, price to US compareto control groups
Groups Obs Mean Std.dev Min Max Growth
Quantity in 100 kgValue in thousand Euro
This study evaluates the impact of advertising on Vietnam's footwear industry using data from the Vietnam Enterprise Survey, collected by the Vietnam General Statistics Office (GSO) and its provincial branches The analysis focuses on panel data from the years 2004 and 2006, featuring 91,755 observations in 2004 and 131,347 firms in 2006, including 329 footwear firms in 2004 and 369 instances of treatment in 2006 To enhance robustness, data from 2003 and 2007 were also included A challenge arose in merging the 2006 and 2007 data with the 2004 and 2005 datasets due to changes in Vietnam's business classification system implemented in 2006, necessitating the concordance of codes across these years.
Table 3.4 Descriptive Statistics of Vietnam Firms Characteristics
Year Variables Groups No.obs Mean Std.dev Minimum Maximum
Price in Euros over kg
Our survey focuses exclusively on apparel products as the control group and footwear as the treatment group, utilizing the SITC 4-digit classification Footwear serves as a proxy for firms under AD, while we enhance the control group to include textile, paper, and wooden products for robustness Table 4 provides a comprehensive data description of the variables analyzed, detailing the mean, standard deviation, maximum, minimum, and number of observations for each group across the years 2003, 2004, 2006, and 2007 Revenue and payroll are reported in millions of Vietnam Dong, with capital also measured in millions of Vietnam Dong The female ratio is calculated as the fraction of female employees over total employees, and firm size represents labor force size The analysis includes three groups: one treatment (footwear firms) and two controls (apparel and W.P.T, which stands for wood, paper, and textile products).
3.2.3 US Import 10-digit (HTS10) Data
US import data, specifically the 10-digit Harmonized Tariff Schedule (HTS10) information, is sourced from the US International Trade Commission This data encompasses all imports and exports between the US and other countries, categorized at the most granular level Values are presented in thousands of US dollars, while quantities vary by product line, with footwear measured in dozens or kilograms Average prices are calculated in USD per unit Notably, the average export value of Vietnamese footwear has surged by 629.6%, with a volume increase of 422.8% Similarly, Vietnamese apparel shows significant growth, with values and volumes rising by 657.8% and 483%, respectively Additionally, there is a notable increase in India's footwear exports.
Vietnam's footwear market shows significant growth, with price increases of 66.9%, while Thailand, Indonesia, India, and Vietnam's apparel also experience rising trends at 52.7%, 75.2%, 81.9%, and 70.4%, respectively In contrast, Indonesia and Thailand face declines, reporting a 35% drop in value and a 30% decrease in quantity for each country.
Table 3.5.Vietnam footwear export value, volume, price to US compareto control groups
Groups Obs Mean Std.dev Min Max Growth
Quantity in 1000 UnitsValue in thousand USD
Price in USD over Unit
Model Estimation and Research Findings
Double Difference Approach
This research utilizes the Double Difference (DD) method to analyze the effects of antidumping (AD) measures on the footwear industry By comparing treatment and control groups before and after the policy intervention, the study distinguishes between ex-ante and ex-post periods using the variable YEAR_t, which denotes the respective timeframes This approach accounts for year fixed effects, encompassing macroeconomic factors such as exchange rates and inflation The outcomes of interest, including revenue, profit, cost, labor payroll, and firm size, are represented by YTt for the treatment group and YCt for the control group in year t As highlighted by Khandker et al (2010), the impact of antidumping measures can be effectively assessed through this methodology.
DD = E(YT1 – YC1|T1 =1) – E(YT0- YC0|T0 = 0) (4.1)
Where T1 =1 stands for “named” AD group after intervention while T0 =0 represents for “non-named” AD group This DD assumes unobserved heterogeneity is time-invariant.
The difference-in-differences (DD) approach, represented by E(YT0 - YC0 | T0 = 0), improves upon the before-after method by effectively removing time-invariant biases While this assumption is less stringent than the conditional exogeneity assumption, it provides a more robust framework for analysis.
Yit = α + β*Ti1*YEAR_t + � ∗ �i1+ � ∗ ����_t + εit (4.2) Substitute T1 = 1 and T1 = 0 into (4.2),
As such, equation (4.1) could be equal to
In the Difference-in-Differences (DD) approach, DD is defined as the combination of YEAR_t and the treated group, where the coefficient β represents the policy effect on individual i in group T during year t A crucial assumption of DD methods is that Cov(Ti1*YEAR_t, εit) = 0, indicating that bias remains time-invariant To uphold this assumption, the control group must share common characteristics with the treatment group Furthermore, the Difference-in-Differences model can be adapted, as discussed by Brambilla et al (2012).
The model LnYit = α + Ai + β*Ti*YEAR_t + δ*xi + γ*lnxit + εit incorporates individual characteristics, represented by the vector xit, which includes factors such as firm age, size, capital per employee, and location, all of which vary over time By integrating individual fixed effects (Ai) and these characteristics, the model effectively controls for observed biases, enhancing the accuracy of the analysis.
An illustration of DD method could be seen as figure 11 We reveal case of Appling
The DD method is utilized to assess the impact of Antidumping Duty (AD) on the revenue of Vietnamese footwear firms The analysis divides the study into two distinct periods: before and after the intervention Initially, the revenue for Vietnamese footwear firms is denoted as YiA, while the control group's revenue is represented as YiB Following the implementation of the AD, the revenues are recorded as YiA2 for footwear and YiB1 for the control group.
Figure 4.1 Difference in Difference Method
The expected revenue (YiA1) of a footwear enterprise without anti-dumping (AD) measures is a theoretical concept, as it has never been observed in practice The DD approach posits that the difference in revenues (YiA – YiB) is equivalent to the difference in expected revenues (YiA1 – YiB1) Consequently, the impact of EU anti-dumping measures on the revenue of Vietnamese footwear firms can be expressed as (YiA2 – YiA) – (YiB1 - YiB) This methodology can similarly be applied to assess the effects of AD on various aspects of footwear firms, including profit, labor payroll, and overall firm size.
Double Difference in Multiple Years
Model (4.4) could be generalized for multiple years as described in (Blanchflower,Oswald, & Sanfey, 1996)
LnYit = α + Ai + β*Ti*YEAR_t+�_ � ∗ ����_t + �*lnxit+ εit (4.5)
YEAR_t represents year fixed effects, showing a slight difference in value between equations (4.4) and (4.5) Unlike (4.4), which has binary values of 0 and 1, (4.5) encompasses a range of years with available data We apply first differences to both sides of the equation, eliminating time-invariant heterogeneity, which results in (Ai – Ai) equaling zero Additionally, Xit accounts for the control of time-variant individual characteristics.
The analysis incorporates year fixed effects to estimate the influence of policy on the outcome variable Yit We adapt our model by modifying the covariate to β*Ti*AfterAD_t, where AfterAD_t indicates the period before and after the policy intervention with values of 0 and 1 Additionally, we include country fixed effects to account for variations across different nations To facilitate a comparison of Vietnam's footwear industry with other countries, the model is further refined by incorporating Ai as a control for industry fixed effects.
LnYst = α +Ai+ Countrys+ ω_t*YEAR_t + β * Countrys* AfterAD_t + δ*Xst + εst (4.6)
Besides that, to make our results stronger, we also run regression for only VN industries as described in 4.6b Footwear is treatment; other industries are control.
LnYit = α + Ai+ ω_t*YEAR_t + β * Industry_i * AfterAD_t + δ*X_it + εit(4.6b)
Industry_i is dummy variable control industry which is equal to 1 if industry is footwear;otherwise it takes value of zero.
Estimation Impact of EU Antidumping on Vietnam Footwear Export
To evaluate influence of EU AD on Vietnam footwear Export at industry level, we employ equation4.6 with a slight adjustment To assess AD effects on value import 4.7 is modified as
LogValueist= α + Ai+ Vn+ β*ddVN+ �_ � ∗ ����_t +εit (4.7)
The LogValueist represents the logarithm of value imports to the EU from industry i in country s during year t, with Vn indicating country-fixed effects, assigning a value of one for Vietnam and zero for other countries The binary variable ddVN captures the interaction between industry i and country s (Tis) alongside AfterAD_t, where Tis equals one if the industry is included in the EU's "named" anti-dumping (AD) list and originates from Vietnam, otherwise it is zero YEAR_t, AfterAD_t, and εit are defined as per equations 4.7 In equations 4.8, 4.9, and 4.10, ddVN functions as the difference-in-differences (DD) variable, while Ai accounts for industry time-invariant fixed effects A parallel equation to 4.8 is utilized to assess the impact of AD on Vietnam's export quantities to the EU.
LogQuantiyist= α + Ai + Vn+ β*ddVN+ �_ � ∗ ����_t + εit (4.8)
LogQuantityist signifies the quantity of imports into the EU from country s for industry i in year t This study further examines the influence of EU anti-dumping (AD) measures on the pricing of AD footwear imported from Vietnam, as illustrated in equation 4.10.
LogPriceist= α + Ai + Vn+ β*ddVN+ �_ � ∗ ����_t+ εit(4.9)
LogPriceist stands for log of price import to EU in industry i from country s in year t In short, a summary for variables and their definitions was illustrated in table1 of Appendix A
Figure 4.2 presents data on the export of Vietnamese AD footwear and apparel from Thailand, Indonesia, and India to the EU The analysis focuses on the trends in value, volume, and price from 1999 to 2004 to establish a control group, with India emerging as the closest comparator to Vietnamese footwear during this period Notably, the European Commission used India as a reference for assessing claims regarding Vietnamese footwear pricing Since 2007, Thailand's footwear exports have declined, likely due to political instability following the 2006 coup Additionally, from 2001 to 2004, there was a marked decrease in apparel value for Indonesia, Thailand, and Vietnam compared to the control group Indonesia experienced a rapid increase in quantity from 2000 to 2002, followed by a sharp decline from 2002 to 2004, while Thailand continued to see significant decreases post-2007 Furthermore, both Indonesia and Vietnam's apparel prices have notably dropped during this period.
Between 2000 and 2004, India serves as the most suitable control group for our analysis To ensure robustness, we also conduct regressions on the apparel sectors of Indonesia, Thailand, and Vietnam to observe how varying control groups may yield different results aligned with their respective trends Furthermore, we analyze the period from 2003 to 2007 to assess the short-term impact of Anti-Dumping (AD) measures These approaches aim to validate the Difference-in-Differences (DID) methodology and strengthen our research conclusions.
Figure 4.2 Trend of export value, quantity, price of Vietnam AD Footwear and control groups to EU
The results presented in Tables 4.1, 4.2, and 4.3 demonstrate the impact of Anti-Dumping (AD) measures on the export value, quantity, and price of Vietnamese footwear to the EU, utilizing equations 4.7, 4.8, and 4.9 All regressions exhibit an F-test probability of less than 1%, allowing us to reject the null hypothesis that all coefficients are equal to zero The R-squared values range from 0.0816 to 0.6133, with the exception of 0.0386 in column (8) of Table 4.3, indicating the percentage of variations in the logarithm of value, quantity, and price explained by the independent variables The intercept values, representing the dependent variables when both Vn and ddVN are zero, are statistically significant at the 1% level across all three tables Additionally, the models account for year-fixed effects and product-fixed effects in the regressions.
The analysis of the EU tax's influence on Vietnam's export value reveals significant effects, as indicated by the covariate ddVN, with statistically convincing estimations at the 1% level across all columns except for column (3) This trend may be attributed to Thailand's political crisis in 2007 Focusing on the period from 2003 to 2007, column (7) shows a notable 34.6% reduction, also statistically significant at the 1% level Additionally, in the context of India, columns 1 and 5 demonstrate that the AD leads to a substantial decline of 105.6% and 88.8%, respectively, corresponding to the year 1999.
Between 2003 and 2007, the impact of antidumping (AD) measures on Indonesia and Vietnam's apparel industries was significant, with control group estimates showing 52.8% for Indonesia and 69.5% for Vietnam, escalating to 88.6% and 97.4% respectively This increase is attributed to Vietnam's remarkable growth and Indonesia's strong recovery compared to the average trend from 1999 to 2011 However, both countries exhibited lower estimates than India, likely due to declines in the 2001-2003 period The variable for Vietnam accounts for country-specific fixed effects, revealing that Vietnam's exports were 87.8% and 181.5% higher than those of Indonesia and Thailand from 1999 to 2011 When focusing on the 2003-2007 timeframe, Indonesia and Thailand were found to be 143.3% and 209.7% lower than Vietnam, respectively Notably, there was no statistically significant difference between India and Vietnam in columns (1) and (4).
Table 4.1.Vietnam AD footwear export value to EU compares to control groups
Year FE Yes Yes Yes Yes Yes Yes Yes Yes
Product FE Yes Yes Yes Yes Yes Yes Yes Yes
R-Square 0.2213 0.1834 0.4558 0.1021 0.0928 0.4075 0.5886 0.1032 Note: *, **, *** indicate 1%, 5%, and 10% significance level, respectively (* p