DOI QR코드

DOI QR Code

Determinants of the World's Rice Trade: The Role of Trade Costs

  • 투고 : 2022.02.10
  • 심사 : 2022.05.10
  • 발행 : 2022.05.30

초록

We investigate the impacts of tariff, tariff-rate quota, conformity assessment, and traceability requirements set by the rice importers using panel data of 17 rice exporters exporting to 119 countries in two years 2015 and 2018, using both Ordinary least square (OLS) and Poisson pseudo maximum likelihood (PPLM) methods. Results from our gravity models strongly indicate that tariff and tariff rate quota remain significantly detrimental to the world's rice exporters because these measures place downward pressure on the rice exporters' prices and the importers' import quantity, creating barriers to market participation. Our study also provides strong evidence about the role of origin certificates in allowing rice exporters to gain access to foreign markets. Meanwhile, regulatory standards such as traceability requirements and logistical and distributional requirements imposed by the rice importers are found to be major obstacles to rice importation from less-developed countries. Our study provides insights into tariff and non-tariff barriers existing in the global rice market, which is likely to assist policymakers operating in developing countries to help shape their policies and bolster rice competitiveness.

키워드

1. Introduction

Non-tariff measures (NTMs) refer to policy measures to regulate market access and/ or ensure that imported products are aligned with public issues/concerns. This includes food safety standards, pest control, and animal disease control UNCTAD (2019). The UNCTAD categorizes these NTM measures into 16 chapters which are grouped into two types, i.e. technical and non-technical measures. Technical measures include sanitary and phytosanitary measures, technical barriers to trade (TBT), and inspection before shipment, while non-technical measures consist of the measures that would intentionally affect trade, such as contingent trade protective measures, import control, or quota measures.

Studies on the impacts of NTMs on trade at the country level show mixed impacts. These inconsistent results are attributed to, as pointed out by Santeramo and Lamonaca (2019) who reviewed more than 140 empirical NTM-relevant studies, the variation in types of NTM measures investigated. This includes various proxies used for NTMs, levels of detail of studies, various methodologies, and publication processes across the studies reviewed.

Agricultural, food, and animal products are subject to NTM measures more frequently than manufacturing products (Cadot et al., 2018; Ing et al., 2015), and this trend has been on the rise in agri-food sectors (Devadason, 2020; ESCAP, 2019; KIEP, 2020; Li & Yu, 2020). In regards to sector specific analysis, a number of previous studies investigate the impacts of NTM measures on the seafood exports from Africa to the EU (Kareem, 2016), the impacts of chemical standards on fish imports into the 15 EU members, Japan, and North America (Tran et al., 2013), the determinants of the international seafood industry (Natale et al., 2015). Some other NTM-related and sector-specific studies also include meat products exported by the USA (Schlueter et al., 2009), exports of corn seeds from the USA (Jayasinghe et al., 2010), and exports of groundnuts from Africa to the EU (Xiong & Beghin, 2011). However, to the best of the authors’ knowledge, there is a lack of in-depth analysis of the impacts of NTMs on the rice sector. Rice becomes the focus of this study because of three reasons.

First, the rice trade has been heavily regulated by tariffs and non-tariff measures. In addition, it has long been known as a thinly traded market, with only 6.94% of the rice supply being traded internationally since 2010 (Durand-Morat & Bairagi, 2021). This fact could be attributed to food security concerns in countries such as Malaysia, Indonesia, Vietnam, China, South Korea, and Japan (Ali et al., 2019; Greenville, 2018). In addition to high tariffs on rice – with the average regional rate close to 25% among ASEAN member countries - almost all member states imposed non tariff measures (NTMs) on rice, with over 400 non-tariff measures in 2019 as opposed to only 94 in 2000. Most of these measures were related to SPS measures and were implemented to protect plants, animals, and human health (Greenville, 2018). Additionally, the net rice importers in ASEAN, namely Indonesia, the Philippines, and Malaysia, also typically employed trade-controlling NTM measures, e.g. licensing and/or bans on private importation of rice (outside nominated monopoly importers) to control rice imports. In Indonesia, for example, Marks (2017) estimates that in 2015 tariff and non-tariff measures levied on the rice imports into Indonesia resulted in an effective tariff – as measured by the nominal rate of assistance – of around 67%, which is mainly attributed to quantitative limits created by licensing. Table 1 presents a summary of the rice policies of several key players in the world’s rice market.

Table 1: Rice Policies of the World’s Important Countries in the World’s Rice Market

*Not mentioned. Source: FAO (2018) and FAO (2015).

Although rice has undoubtedly been the most heavily regulated agricultural commodity, in-depth analysis of these trade-controlling measures on rice appears neglected. Therefore, this study aims to fill this gap by investigating the impacts of tariff, tariff-rate-quota, conformity assessment, and traceability measures, such as testing, certification, or inspection, on the world’s rice market. Our paper employs the most sophisticated NTM database, which was made available by UNCTAD (2017), instead of employing NTM notifications to the World Trade Organization (WTO), under obligations required by the Agreement on the Application of Sanitary and Phytosanitary Measures or the Agreement on Technical Barriers to Trade or NTM data as specific concerns by WTO members. This new NTM dataset allows us to simultaneously examine the impacts of tariff and non-tariff measures imposed on the world’s rice market. Conformity assessment measures are selected because, as pointed out by Keiichiro et al. (2015) and Webb et al. (2018), such measures result in an increase in compliance cost for producers and/or exporters, but in the future, this cost can be reduced by mutual recognition of standards (Cadot et al., 2015; Crivelli & Groeschl, 2016; Keiichiro et al., 2015).

Second, although the literature provides profound statistical evidence proving the disproportionate impact of NTM on countries of different levels of economic development, to the best of our knowledge, there is a lack of such empirical evidence found specifically in the world’s rice market. A number of previous studies have proven that the stricter standards imposed by developed countries deter developing countries from exporting their goods to developed markets (Cadot & Gourdon, 2016; Crivelli & Groeschl, 2016; Ferro et al., 2015; Kee et al., 2008; Mahdi et al., 2016; Webb et al., 2018). Webb et al. (2018) analyzed the impacts of NTMs imposed by four developed countries, namely New Zealand, the USA, the EU, and Canada, on their trading partners. Their findings showed that the more measures imposed by such importers related to conformity requirements, i.e. testing, certification, or inspection, the less likely developing countries were willing to export to these countries. We attempt to compare the impacts of selected conformity and traceability measurements on developing rice exporting countries as opposed to their developed counterparts. Together with conformity requirements, which are coded under A82, A83, and A84 under UNCTAD (2017) classification, our analysis also includes A85 measures (traceability requirements). There are two main reasons for this inclusion. As established in the literature, traceability has become an increasingly important component of food safety and quality regulations, management systems, and certification processes. It ensures food quality and builds consumers’ trust by improving transparency in the supply chain, thus improving market entry for agricultural products (Maertens et al., 2012; McCullough et al., 2010). Therefore, we attempt to analyze the impacts of these measures for the purpose of guiding policy decisions of the developing countries that wish to get further access to foreign markets.

The third difference between our study and from preceding ones lies in the methodology that allows us to distinguish the trade-enhancing and/or trade-hindering effects of each selected conformity assessment or traceability measure. This distinction further allows us to provide sound policy recommendations, too.

In brief, by employing new NTM data provided by UNCTAD, this study attempts to address three research questions. First, to what extent do tariff and tariff rate quota measures negatively affect the global rice trade? Second, how do the conformity assessment and traceability measures affect the rice trade? And finally, how are the impacts of these conformity assessments and traceability measures on the rice from developing countries different from the rice from more developed countries?

Our paper is organized into 6 sections. In Section 2, we provide literature reviews followed by our rationales for our choice of estimation method and our description of data used in this research in section 3. The fourth section presents our results, and finally, the last section provides conclusions and policy recommendations.

2. Literature Review

Previous research has firmly established the negative impacts of these trade-deterring measures on the rice trade. (Ali et al., 2019; FAO, 2018; Furuhashi & Gay, 2017; Greenville, 2018; Nguyen, 2020; Nguyen et al., 2020). However, with regard to non-tariff measures, there has been a lack of studies paying particularly in-depth attention to one single commodity. Meanwhile, studies on aggregated agricultural commodities show mixed viewsSPS, and TBT rules, according to Disdier et al. (2008), have no major impact on bilateral trade between OECD members, but they dramatically restrict exports from developing and less developing nations to OECD countries, with the exception of dairy goods, milling products, and rice. Because SPS and TBT regulations are not necessarily intended to have a detrimental impact on trade, these authors advocate for additional technical help to poor nations, allowing them to achieve the level of safety mandated by SPS or TBT measures on the global market. In contrast, Disdie et al. (2008) and Cadot and Gourdon (2016) employed more detailed and sophisticated data from UNCTAD and found that SPS regulations are particularly stringent for food products, thus raising production costs in general. Specifically, their imperial results proved that SPS measures raise the price of animals, vegetables, fats, and oils by 12.9%, 10.3%, and 6.9%, respectively (Grant et al., 2015). These findings are consistent with the findings from Ferro et al. (2015), who also ascertained that more restrictive standards measured by the trade-restrictive index are associated with a lower probability of observing agricultural trade. Also, based on the previous results, Cadot and Gourdon (2016) stated that the imports that comply with SPS regulations have strong quality and signaling effects, thus raising demand for such products. This statement, however, appears questionable as their research employed a unit price-based independent variable rather than quantity. In fact, based on the quantity data on corn seeds, Jayasinghe et al. (2010) found that SPS measures are among the third significant barrier to the US’s exports of corn seeds after tariff and distance.

Despite variations in the methods and source of data used in the three aforementioned studies, authors of these studies agree that SPS and TBT measures add more cost to the exports and exert more trade-impeding impacts on developing countries than within developed countries. This finding is also reinforced by several other studies (Fontagné et al., 2015; Keiichiro et al., 2015; Murina & Nicita, 2017; Xiong & Beghin, 2017). Therefore, Cadot and Gourdon (2016) highlighted the need to enhance coordination and adherence to the established international standards because such initiatives would enable under-developed producers to reduce their compliance costs.

Following the similar methodology that was first introduced by Cadot et al. (2018) and using a more updated NTM dataset by UNCTAD (2017), the recent study by Sung et al. (2020) explored the impacts of structural similarity of non-tariff measures on agricultural trade worldwide and found that while a structural similarity in technical barriers to trade (TBT) between pairs of countries is positively correlated with an increase their bilateral trade, the opposite relationship is found in the case of sanitary and phytosanitary measures SPS. Based on this finding, Sung et al. (2020) also put forward their recommendations for the harmonization of trade regulations which they believe are likely to reduce trade costs for agricultural products, thus boosting agricultural and food trade in the future.

Although studies like Cadot et al. (2018) and Sung et al. (2020) that use aggregated data to support future NTM standard harmonization provide useful policy recommen- dations, they cannot provide sector-specific policy recommendations that are needed for trade negotiations or policy-making processes to help a particular commodity of interest, especially from developing countries, gain better access to strict foreign markets. In contrast, our analysis concentrates on just the world’s rice market and deeply investigates the impacts of conformity measures on the world’s rice trade rather than all SPS measures using the latest NTM data by UNCTAD (2017). We restate the trade-impeding effects of tariffs and tariff-rate quotas on the world’s rice market, and we also provide sufficient evidence for the facilitating role of origin certificates in allowing rice exporters to gain access to foreign markets. Another remarkable result is those traceability standards and logistical and distributional requirements to share information among participants in the rice supply chain are found to be the major barriers against the rice import from less-developed countries.

3. Research Methods and Materials

3.1. Model Specifications

Econometrical estimations of the determinants of bilateral trade have been commonly based on gravity econometric model(s) which was first introduced by Tinbergen (1962). Over time theoretical foundations of the gravity model have been developed, and one of the most grounded theoretical frameworks that have been widely applied in trade studies is the model developed by Anderson and van Wincoop (2003). This model is described below:

\(X_{i j}=\frac{Y_{i} Y_{j}}{Y}\left(\frac{t_{i j}}{\prod_{i} P_{j}}\right)^{1-\rho}\)       (1)

Where Y denotes world GDP. Yi and Yj represent the GDP of countries i (exporter) and countries j (importer), respectively. t ij (one plus the tariff equivalent of overall trade costs) is the cost in country j of importing a product from country i. Typically, previous empirical studies proxy trade costs with bilateral distance along with other dummy variables, namely islands, landlocked countries, and common borders. σ > 1 is the elasticity of substitution. Пi represents i’s outward multilateral resistance terms. Pj represents j’s inward multilateral resistance terms.

As for the multilateral resistance terms (MRTs), there are different proxies for these terms, but as suggested by Anderson and van Wincoop (2003), one convenient way is to replace the multilateral resistance indexes with the importer and exporter dummies or a pair country effects. This method helps avoid the possibility of biased estimates of the impact of distance and other bilateral variables on bilateral trade flows.

In the case of a sectoral study, as demonstrated by Anderson and van Wincoop (2003, 2004), the sectoral gravity equation is depicted as follows.

\(X_{i, j, t}^{k}=\frac{Y_{i, t}^{k} E_{j, t}^{k}}{Y_{t}^{k}}\left[\frac{t_{i, j, t}^{k}}{P_{j, t}^{k} \pi_{i, t}^{k}}\right]^{1-\rho_{k}}\)   (2)

where , Yki,t stands for country-level output/production, and , Eki,t For expenditure, and k represents a commodity. tki,j,t stands for trade cost, which is sector-specific, Pki,t and πki,t are multilateral resistant terms that are also sector-specific.

Xiong and Beghin (2017) introduce a theoretical model approach that could take into account both demand enhancing impacts and trade-cost effects of NTM measures. Based on the theoretical models in Xiong and Beghin (2017) and the estimation model suggested by Cadot et al. (2018), our empirical model for the price is specified as follows.

\(\begin{aligned} \text { In } p_{i j k}=& \beta_{1} \operatorname{In}\left(\text { production }_{i k}\right)+\beta_{2} \operatorname{In}\left(\text { consumption }_{j k}\right) \\ &+\beta_{3} \text { AVEtarif }_{i j}+\beta_{4}\left(\text { distance }_{i j}\right)+\beta_{5} \text { Contingent }_{i j} \\ &+\beta_{6} \text { Common language }_{i j}+\beta_{7} \operatorname{In}\left(\text { Totalimport }_{j k}\right) \\ &+\sum_{m} \beta_{8 m} n_{j k m}+\sum_{i} \delta_{i}+\sum_{j} \delta_{j}+\sum_{k} \delta_{k}+u_{i j k} \end{aligned}\)   (3)

Where pijk denotes for import price of good k from an exporter i to an importer j δi δj δk are importer, exporter, and product fixed effect, njkm refers to the number of SPS measures under the MAST’S one letter + two-digit category imposed by importers j on product k.

The volume equation to estimate with PPML is as follows:

\(\begin{aligned} q_{i j k}=& \beta_{1} \operatorname{In}\left(\text { production }_{i k}\right)+\beta_{2} \operatorname{In}\left(\text { consumption }_{j k}\right) \\ &+\beta_{3} \text { AVEtarif }_{i j}+\beta_{4}\left(\text { distance }_{i j}\right)+\beta_{5} \text { Contingent }_{i j} \\ &\left.+\beta_{6} \text { Common language }_{i j}+\beta_{7} \operatorname{In} \text { (Totalimport } \text { Tok }_{j k}\right) \\ &+\sum_{m} \beta_{8 m} n_{j k m}+\sum_{i} \delta_{i}+\sum_{j} \delta_{j}+\sum_{k} \delta_{k}+u_{i j k} \end{aligned}\)  (4)

It is generally agreed in the literature that trade dat abased studies often have to deal with the typical issues of heteroscedasticity and zero-trade observations. WTO (2012, p. 114) highlights that “sectoral trade flows are likely to be more heterogeneous than aggregate ones − in which sectoral idiosyncrasies are averaged out − so outliers and heteroscedasticity should be handled with special care.” Likewise, after reviewing about 150 previous studies on non-tariff barriers, Santeramo and Lamonaca (2019) highlight the issues of inclusion of multilateral resistance terms and treatment of zero trade flows in NTM-related studies because the greater chance of observing statistically significant depends on the robust methodological appro- aches, inclusion of multilateral resistance terms and treatment of zero trade flow. In line with Santeramo and Lamonaca (2019) and Will (2020) identified three main problems faced by the research using gravity models. These problems include limited-dependent variable bias, a combination of heteroscedasticity and nonlinearity, and zero trade issues. With regards to the latter issue, there are three possibilities of zero trade. First, zeros may also be the result of rounding errors; second, zeros can simply be missing observations that are wrongly recorded as zeros; and third, zeros may be the result of firms’ decision not to trade

Given that zero trade and heteroscedasticity can be the main sources of biased estimation, our next section presents our rationale for our approach to treating zero trade and our choice of estimation approach for our analysis.

3.2. Research Method

A large number of previous studies on the determinants of bilateral trade flow with disaggregated data commonly employed OLS, Heckman estimation, Poisson pseudo maximum likelihood model, and its variant models (zero- inflated models). Other methods include the fixed effect model, random effect model, and Tobit model. The Tobit model is one of the common approaches used in firm level studies, which goes beyond the scope of this study. Fixed effect and random effect models are not employed in this study either due to the following reasons. First, one dependent variable used in this study is bilateral trade which takes continuous value starting from 0 to thousands, and so do the remaining variables. As such, simple quartile plots prior to regression usually support the use of log-linear rather than linear models. However, as the log of zero means unidentified, zero value is dropped out before regression when these models are applied in a statistical program like Stata. Because one of the aims of the study is to capture the information conveyed from zero trade quantity, these three methods are not aligned with this purpose, and therefore the OLS model is employed in the model of price unit and is not employed in the model of quantity.

The random effects model can take into account unobserved country-specific factors that may affect trade, but this model assumes the unobserved time-invariant component must not be correlated with one of the control variables. If so, the estimates are biased and not consistent. An alternative for the random-effects model is the fixed effects estimation. However, this estimation eliminates invariant time variables, such as geographical distance.

There exists a consensus among the gravity-based literature reviews that Pseudo Poisson Maximum Likelihood estimation (PPML) in trade studies has been one of the most commonly used methodologies to deal with zero trade and heteroskedastic issues when the data used does not have an overdispersion issue (Silva & Tenreyro, 2006). Recent researches using this model include the ones by Dong and Truong (2022), Tran et al. (2013), Sung et al. (2020), and Webb et al. (2018, 2020).

3.3. Data Used in the Model

We apply the method suggested by Cadot et al. (2018) employing dependent variables in the form of both trade volume and unit price instead of trade value. When disaggregated NTM data on rice is available, we employ panel data from 17 major rice exporting nations to 119 major importing countries between 2015 and 2018. These 17 countries include Argentina, Australia, Belgium, Brazil, China, Spain, India, Cambodia, the Kingdom of the Netherlands, Pakistan, Paraguay, Thailand, Turkey, Uruguay, The United States of America, and Vietnam. Although only 17 main rice exporters are included in the model, they account for a majority of the world’s paddy rice production and export of processed rice, namely more than 90% in 2015 and 2018 with five major countries accounting for 75% of the total volume of exports in 2017–2019 (Ali et al., 2019). As the world’s rice trade is “thin, ” with the supply of the world’s rice exports mainly coming from these rice exporters (USDA Economic Research Service, 2021), we believe our selection of 17 rice exporting countries in the data will help us avoid the issue of downward bias due to excessive zero observations.

Trade, rice consumption, and production data are all sourced from the FAO database (FAO, 2021). We do not use GDP as a proxy for demand because the use of GDP value as a control variable is insignificant in much research on a single commodity (WTO, 2012). Tariff data is retrieved from World Integrated Trade Solution (2021). When available, the bilateral applied tariffs (AHS) are chosen; otherwise, the most-favored-nations tariffs (MFN) are selected instead. When both types of tariffs are not available for a given year, the tariff levied in the previous year is selected instead. Because the gravity model works with logarithms and the logarithm of a zero tariff is undefined, we transform the tariffs using this formula = In(x + (x2 + 1)0.5). This transformation allows us to have the same results as using logarithms for non-zero values without making ad hoc adjustments to zero trade values (Gibson et al., 2017). Because preferential tariffs already reflect regional free trade, we do not include a free trade agreement FTA dummy in the model. When rice is excluded from regional trade integration, the tariffs applied on rice imports from RTA members and non-RTA members are usually the same.

NTM data is retrieved from UNCTAD (2017). This database was updated and relaunched in 2017. NTMs are categorized into 16 chapters (Table 2). This study is limited to studying the impacts of conformity assessment measures (A8) under chapter A and the impacts of tariff-rate quotas under chapter E of the MAST (UNCTAD, 2019). We control for importer, exporter, and year through dummy variables. Geographical distance, contiguity, and common language are all sourced from the CEPII database (Mayer & Zignago, 2011).

Table 2: Summary of Descriptive Statistics

4. Results and Discussion

As shown in Table 3, in Model 1, geographical distance and production output of paddy rice significantly affect rice price. As expected, rice price becomes relatively higher when market destinations are far away from market origins. In contrast, unit prices became relatively lower when paddy rice production in exporting countries increased. As far as the impacts of tariff-rate quota measures are concerned, E211 is found to place downward pressure on rice prices at a significant level of 10%.

Table 3: Estimated Effects of Tariff and Tariff-Rate Quota Measures on the Unit Price of Rice

Note: *** and ** and * indicates 1%, 5% and 10% level of significance.

The results of Model 2 in Table 3 show that, as anticipated, favorable factors for rice trading quantity include paddy production in exporting countries, demand for rice consumption in importing countries, total imports of rice from importing countries, and contingency to market destinations while unfavorable factors include tariff and distance. The negative signs of tariff and distance reaffirm the literature that these two factors remain the major hurdles to the world’s rice trade. Surprisingly, having a common language is found to be unfavorable rather than supportive among rice traders, at a 10%, significant level. This unexpected finding could be the result of their similarity in food culture and agricultural production. In consequence, these countries become direct competitors in the rice market rather than trading partners because they might export similar types (s) of rice.

As far as the impacts of quota-related measures on rice trade quantity are concerned, as shown in the last column of Table 3, the presence of E119 (licensing for the protection of public health) and E611 (WTO-bound tariff rate quotas) measures are found to cause rice trading quantity to decline at the significance level of 10% and 5%, respectively. Another key feature of Table 3 worth mentioning is although E211 (global allocation of quota) significantly affects rice price in Model 1, this effect becomes insignificant in Model 2. In brief, the negative estimates of E119 (import license), E611 (tariff-rate quotas), tariff, and transportation cost (represented by geographical distance) reinforce the fact that these factors remain detrimental trade barriers to the global rice trade.

As shown in Table 4, the estimated coefficients of the explanatory variables in Table 4, remain relatively consistent in both sign and significance with the findings presented in Table 3.

Table 4: Impacts of Conformity Assessment and Traceability Measures on Rice Price and Import Quantity

Note: *** and ** and * indicates 1%, 5% and 10% level of significance.

Because one focus of this study is to quantify the impacts of conformity assessment and traceability measures, our findings suggest that while A84, A852, and A853 are found to hinder trading quantity in Model 4, A851 is found supportive. This can be attributed to the exporters’ ability to provide evidence of the product’s origin which could entrust consumers with the rice quality, thus increasing the demand. This finding is consistent with the literature (Ferro et al., 2015; Kee et al., 2008; Latip et al., 2021; My et al., 2018; Pham, 2020; Webb et al., 2018). My et al. (2018) stated that Vietnamese consumers are willing to pay a 9% price premium for certified sustainably-produced rice (Okpiaifo et al., 2020). This premium has gradually increased by 33% when incremental levels of information on certification and traceability are provided. Likewise, Glory et al. (2020) highlighted that the authentication method protects premium rice brands, combats commercial fraud, and quickly locates the origin of contaminated rice to add verifiable food safety measures for Nigerian rice consumers. Food safety is also identified as the most important sustainability attribute for rice consumers in Nigeria (Glory et al., 2020) as well as South and Southeast Asian countries (Bairagi et al., 2021). Agricultural products from Europe with clear certifications of origin are associated with a high export value (European Commission, 2021). European agricultural producers can sell products at a higher value to consumers looking for authentic regional products thanks to the European Geographical Indications on their agricultural products (European Commission, 2021).

As far as the impacts of A852 and A853 under the traceability category are concerned, our finding clearly shows that import quantity reduces drastically compared to those of importers without these measures imposed. These regulations require rice suppliers to disclose information about their processing, logistical and distributional process among importing partners, partially to reduce information asymmetries, and increase product quality management and transparency among the rice chain participants. On the one hand, stringent food safety and traceability requirements cause a new set of transaction costs for small-scale producers faced with inadequate capital investment and poor public infrastructure (McCullough et al., 2010). On the other hand, rice exporters who already have implemented a traceability system in their home countries own competitive advantage over their lesser-developed competitors. In Thailand, for example, rice supply chain management transparency has enabled Thai rice manufacturers and exporters to gain more access to overseas markets (Kittipanya-ngam & Tan, 2020).

To investigate the impacts of A852, A853, and A859 on the rice exporters of different levels of economic development, we disaggregate the data used for Model 4, and then we run two separate models. One model includes only data from less-developed exporters (exporting countries with lower middle income and below) (Model 5). This subgroup includes four countries, i.e. India, Cambodia, Pakistan, and Vietnam, while the latter subgroup includes the remaining countries (higher middle-income countries and high-income countries). Outputs of Model 5 (less-developed countries) and Model 6 (developed countries) are also displayed in Table 4.

The coefficients of production, geographical distance, total demand for rice import, common language, the origin of the products, and quarantine requirement show consistency with the findings as shown in Model 4 of Table 4. As expected, while A852 (processing history) and A853 (distribution and locations of products after delivery) are found to have a negative effect in the Model 4 at a 10% significant level, these factors are also found to be negative in Model 5 and 6. However, one interesting finding from Table 4 is that although the coefficient of A859 (traceability requirements) is found insignificant in Model 4 and Model 5, it is positive and significant at a 5% level in Model 6. This evidence suggests that requirements on traceability imposed by the importers may work to the advantage of the rice from developed countries. In other words, these exporters can better deal with these market entry requirements compared to their less experienced peers.

This aforementioned interesting finding is consistent with several previous studies. Webb et al. (2018) analyzed the impacts of NTMs imposed by four developed countries, namely New Zealand, the USA, the EU, and Canada, on their trading partners and found that more measures that were related to conformity requirements, such as testing, certification or inspection were imposed by such importers, a smaller number of exporting countries selling their products to these four developed countries. This finding is also aligned with the agricultural trade literature stating that exports from developing countries are particularly deterred by stricter standards imposed by developed countries (Ferro et al., 2015; Kee et al., 2008; Mahdi et al., 2016).

5. Conclusion and Recommendations

5.1. Conclusion

Using the NTM data from UN COMTRADE, this study attempts to quantify the impacts of tariff, tariff rate quota, and multiple conformity assessment and traceability requirements under A81, A82, A83, A84, and A85 categories of the MAST, on the world’s rice market. The study provides strong evidence that tariff and tariff rate quota measures remain the significantly detrimental factors affecting the world’s rice exporters because these measures place downward pressure on the exporters’ prices and the importers’ trade volume.

In ASEAN member states, in particular, such highly protected rice trade policies have long been subject to criticism because these policies result in the misallocation of resources. Therefore, as predicted by Greenville (2018), when further regional rice integration is projected to occur in line with greater reform in the agro-food sector across the ASEAN member states, this initiative might lead to several benefits for the ASEAN region as a whole. These benefits would enable ASEAN countries to leverage their comparative advantages in agro-food production and increase economic opportunities for rice farmers so that they can sustain their farming business and increase national food security overall. Advocacies of further tariff reduction in the ASEAN’s rice sector are also presented in the recent publications (Jati & Premaratne, 2017; Valera et al., 2021).

The second aim of this study is to investigate the effects of conformity assessment and traceability measures on the world’s rice trade. Regarding this matter, our study makes two valuable contributions: First, we provide statistical evidence of the negative effect of inspection requirements on all rice exporting countries. With regards to traceability requirements, while origin-related requirements are found to enhance rice trade, logistical and distributional requirements among participants in the rice supply chain are proved to be barriers. More interestingly, traceability requirements under A859 are found to have disproportionate impacts on less developed countries than on developed countries.

To enhance their competitiveness, rice exporters from developing countries must form a collective effort with the government leading the way. Take Vietnam as an example: Rice is the second largest commodity after seafood products because the brands of the majority of its agricultural exports, with rice being the second largest, are often invisible to their users or end-consumers (Pham et al., 2019). With that being said, Pham et al. (2019) recommended Vietnamese policymakers consider a structural shift from primary agricultural products to differentiated and value-added ones. Another suggestion to help increase its rice competitiveness also includes introducing regulations to require rice producers to abide by certain production practices that have been regionally/internationally recognized. This suggestion is also emphasized in the latest publication of Connor et al. (2022).

5.2. Limitations and Recommendations for Future Research

Estimation methodology in trade studies has long been an ongoing debate in the relevant literature. Tran et al. (2013) compared their results from PPML and other alternative estimators and find that the Heckman sample selection and zero-inflated negative binomial (ZINB) models provide the most reliable parameter estimates based on the statistical tests, the magnitude of coefficients, and economic implications. However, later research provided counter-arguments against the findings by Tran et al. (2013) and strongly advocated for the PPML model as the workhorse gravity estimator. For example, Martin and Pham (2019) apply a range of estimators to the SST empirical dataset on international trade and find very little difference between the PPML estimator and its truncated version. The two authors also argue that the results from the truncated PPML estimator are identical to those from the zero-inflated Poisson estimator. Likewise, using two different simulation experiments to compare PPML with its alternative estimators, Will (2020) found that while the result from s Silva and Tenreyro’s (2006) experimental design in their study shows that alternative estimators have a lower bias than PPML, the results from Monte Carlo design, which replicates the much-higher real-world frequency of predicted values near zero, shows the opposite. With all that being said, future research in the global rice trade would be more reliable if it can manage to incorporate the aforementioned estimation methods for robustness check.

참고문헌

  1. Ali, R. B., Alvaro, D.-M., Wailes, E. J., & Luckstead, J. (2019). Self-sufficiency and international trade policy strategies in the Malaysian rice sector: Approaches to food security using spatial partial equilibrium analysis. Asian Journal of Agriculture and Development, 16(1), 25-52. https://doi.org/10.22004/ag.econ.289826
  2. Anderson, J. E., & van Wincoop, E. (2003). Gravity with gravitas: A solution to the border puzzle. American Economic Review, 93(1), 170-192. https://doi.org/10.1257/000282803321455214
  3. Anderson, J. E., & van Wincoop, E. (2004). Trade costs. Journal of Economic Literature, 42(3), 691-751. https://doi.org/10.1257/0022051042177649
  4. Bairagi, S., Gustafson, C. R., Custodio, M. C., Ynion, J., & Demont, M. (2021). What drives consumer demand for packaged rice? Evidence from South and Southeast Asia. Food Control, 11, 129. https://doi.org/10.1016/j.foodcont.2021.108261
  5. Cadot, O., & Gourdon, J. (2016). Non-tariff measures, preferential trade agreements, and prices: New evidence. Review of World Economics, 152(2), 227-249. https://doi.org/10.1007/s10290-015-0242-9
  6. Cadot, O., Gourdon, J., & Tongeren, F. V. (2018). Estimating ad Valorem equivalents of non-tariff measures: Combining price-based and quantity-based approaches. Paris: OECD Publishing. https://doi.org/10.1787/18166873
  7. Connor, M., Cuong, O. Q., Demont, M., Sander, B. O., & Nelson, K. (2022). The influence of climate change knowledge on consumer valuation of sustainably produced rice in Vietnam. Sustainable Production and Consumption, 31, 1-12. https://doi.org/10.1016/j.spc.2022.01.034
  8. Crivelli, P., & Groeschl, J. (2016). The impact of sanitary and phytosanitary measures on market entry and trade flows. World Economy, 39(3), 444-473. https://doi.org/10.1111/twec.12283
  9. Devadason, E. S. (2020). New protectionism. In Journal of Asia-Pacific Business, 21(1), 57-76. https://doi.org/10.1080/10599231.2020.1708232
  10. Disdier, A. C., Fontagne, L., & Mimouni, M. (2008). The impact of regulations on agricultural trade: Evidence from the SPS and TBT agreements. American Journal of Agricultural Economics, 90(2), 336-350. https://doi.org/10.1111/j.1467-8276.2007.01127.x
  11. Dong, C. V., & Truong, H. Q. (2022). Determinants and potential of seafood trade: Evidence from a transitional economy. Foreign Trade Review, 54, 73. https://doi.org/10.1177/00157325221077003
  12. Durand-Morat, A., & Bairagi, S. (2021). International rice outlook: International rice baseline projections 2020-2030. https://scholarworks.uark.edu/aaesrb/49.
  13. El-Enbaby, H., Hendy, R., & Zaki, C. (2016). Do SPS measures matter for margins of trade? Evidence from firm-level data. Applied Economics, 48(21), 1949-1964. https://doi.org/10.1080/00036846.2015.1111987
  14. Economic and Social Commission for Asia and the Pacific. (2019). ASIA-PACIFIC trade and investment report 2019: Navigating non-tariff measures towards sustainable development. https://www.unescap.org/publications/APTIR2019
  15. European Commission. (2021). Study on the economic value of EU quality schemes, geographical indications (GIs), and traditional specialties guaranteed (TSGs). https://data.europa.eu/doi/10.2762/396490
  16. Food and Agriculture Organization (FAO). (2015). The FAO rice market monitor (RMM). https://www.fao.org/economic/RMM
  17. Food and Agriculture Organization (FAO). (2018). The FAO rice market monitor (RMM). https://www.fao.org/economic/RMM
  18. Food and Agriculture Organization (FAO). (2021). Food balances 2014. http://www.fao.org/faostat/en/#data/FBS
  19. Ferro, E., Otsuki, T., & Wilson, J. S. (2015). The effect of product standards on agricultural exports. Food Policy, 50, 68-79. https://doi.org/10.1016/j.foodpol.2014.10.016
  20. Fontagne, L., Orefice, G., Piermartini, R., & Rocha, N. (2015). Product standards and margins of trade: Firm-level evidence. Journal of International Economics, 97(1), 29-44. https://doi.org/10.1016/j.jinteco.2015.04.008
  21. Furuhashi, G., & Gay, H. (2017). Market implications of the integration of Asian rice markets. Paris: OECD Publishing, 108. https://doi.org/10.1787/c81e00c6-en
  22. Gibson, J., Datt, G., Murgai, R., & Ravallion, M. (2017). For India's rural poor, growing towns matter more than growing cities. World Development, 98, 413-429. https://doi.org/10.1016/j.worlddev.2017.05.014
  23. Okpiaifo, G., Durand-Morat, A., West, G. H., Nalley, L. L., Nayga, R. M., & Wailes, E. J. (2020). Consumers' preferences for sustainable rice practices in Nigeria. Global Food Security, 24, 345. https://doi.org/10.1016/j.gfs.2019.100345
  24. Grant, J., Peterson, E., & Ramniceanu, R. (2015). Assessing the impact of SPS regulations on U.S. fresh fruit and vegetable exports. Journal of Agricultural and Resource Economics, 40(1), 144-163.
  25. Greenville, J. (2018). ASEAN rice market integration: Findings from a feasibility study (OECD Food, Agriculture and Fisheries Papers, No. 117). Paris: OECD Publishing. https://doi.org/10.1787/8ca16e31-en
  26. Ing, L. Y., Cadot, O., Anandhika, R., & Urata, S. (2015). Non-tariff measures in ASEAN: A simple proposal. https://www.eria.org/publications/non-tariff-measures-in-asean/
  27. Jati, K., & Premaratne, G. (2017). Analysis of staple food price behavior: Multivariate BEKK-GARCH model. Journal of Asian Finance, Economics, and Business, 4(4), 27-37. https://doi.org/10.13106/JAFEB.2017.VOL4.NO4.27
  28. Jayasinghe, S., Beghin, J. C., & Moschini, G. (2010). Determinants of world demand for U. S. Corn Seeds: The role of trade costs. American Journal of Agricultural Economics, 92(4), 999-1010. https://doi.org/10.1093/ajae/aaq056
  29. Kareem, O. I. (2016). Food safety regulations and fish trade: Evidence from European Union-Africa trade relations. Journal of Commodity Markets, 2(1), 18-25. https://doi.org/10.1016/j.jcomm.2016.07.002
  30. Kee, H. L., Nicita, A., & Olarreaga, M. (2008). Estimating trade restrictiveness indices. Economic Journal, 119(534), 172-199. https://doi.org/10.1111/j.1468-0297.2008.02209.x
  31. Keiichiro, H., Otsuki, T., & Wilson, J. S. (2015). Food safety standards and international trade: The impact on developing countries' export performance. In: Hammoudi, A., Grazia, C., Surry, Y., & Traversac, J. B. (Eds.), Food safety, market organization, trade and development (pp. 151-166). New York: Springer International Publishing. https://doi.org/10.1007/978-3-319-15227-1_8.
  32. KIEP. (2020). Non-tariff measures in the Republic of Korea. ERIA: Economic Research Institute for ASEAN and East AsiaEconomic Research Institute for ASEAN and East Asia.
  33. Kittipanya-Ngam, P., & Tan, K. H. (2020). A framework for food supply chain digitalization: Lessons from Thailand. Production Planning and Control, 31(2-3), 158-172. https://doi.org/10.1080/09537287.2019.1631462
  34. Latip, M. S. A., Newaz, F. T., Latip, S. N. N. A., May, R. Y. Y., & Rahman, A. E. A. (2021). The sustainable purchase intention in a new normal of COVID-19: An empirical study in Malaysia. Journal of Asian Finance, Economics, and Business, 8(5), 951-959. https://doi.org/https. https://doi.org/10.13106/JAFEB.2021.VOL8.NO5.0951
  35. Li, M., & Yu, M. (2020). Non-tariff measures in China. ERIA: Economic Research Institute for ASEAN and East AsiaEconomic Research Institute for ASEAN and East Asia.
  36. Maertens, M., Minten, B., & Swinnen, J. (2012). Modern food supply chains and development: Evidence from horticulture export sectors in Sub-Saharan Africa. Development Policy Review, 30(4), 473-497. https://doi.org/10.1111/j.1467-7679.2012.00585.x
  37. Mahdi, G., Julia, G. B., & Robert, S. (2016). Estimating importer-specific ad valorem equivalents of non-tariff measures. http://hdl.handle.net/10419/203994
  38. Marks, S. V. (2017). Non-tariff trade regulations in Indonesia: Nominal and effective rates of protection. Bulletin of Indonesian Economic Studies, 53(3), 333-357. https://doi.org/10.1080/00074918.2017.1298721
  39. Martin, W., & Pham, C. S. (2020). Estimating the gravity model when zero trade flows are frequent and economically determined. Applied Economics, 52(26), 2766-2779. https://doi.org/10.1080/00036846.2019.1687838
  40. Mayer, T., & Zignago, S. (2011). Notes on CEPII's distance measures: The GeoDist database. SSRN Journal, 21, 117. https://doi.org/10.2139/ssrn.1994531
  41. McCullough, E. B., Pingali, P. L., & Stamoulis, K. G. (Eds.). (2010). Small farms and the transformation of food systems: An overview. Netherlands: Wageningen Academic Publisher. https://doi.org/10.3920/978-90-8686-703-5
  42. Murina, M., & Nicita, A. (2017). Trading with conditions: The effect of sanitary and phytosanitary measures on the agricultural exports from low-income countries. World Economy, 40(1), 168-181. https://doi.org/10.1111/twec.12368
  43. My, N. H. D., Demont, M., Van Loo, E. J., de Guia, A., Rutsaert, P., Tuan, T. H., & Verbeke, W. (2018). What is the value of sustainably produced rice? Consumer evidence from experimental auctions in Vietnam. Food Policy, 79, 283-296. https://doi.org/10.1016/j.foodpol.2018.08.004
  44. Natale, F., Borrello, A., & Motova, A. (2015). Analysis of the determinants of international seafood trade using a gravity model. Marine Policy, 60, 98-106. https://doi.org/10.1016/j.marpol.2015.05.016
  45. Nguyen, D. D. (2022). Determinants of Vietnam's rice and coffee exports: Using stochastic frontier gravity model. Journal of Asian Business and Economic Studies, 29(1), 19-34. https://doi.org/10.1108/JABES-05-2020-0054
  46. Nguyen, H. T. M., Do, H., Kay, A., & Kompas, T. (2020). Rice policy in a transitional economy: Balancing the social and political objectives. Food Security, 12(3), 549-566. https://doi.org/10.1007/s12571-019-01005-x
  47. Pham, D. M., Hollweg, C. H., Mtonya, B., Winkler, D. E., & Nguyen, T. (2019). https://openknowledge.worldbank.org/handle/10986/33219. Connecting Value Chains for Trade Competitiveness, World Bank.
  48. Pham, H. C. (2020). Antecedents of organic food products intention and behaviors: Evidence from Vietnam. Journal of Asian Finance, Economics, and Business, 7(11), 429-437. https://doi.org/10.13106/JAFEB.2020.VOL7.NO11.429
  49. Santeramo, F. G., & Lamonaca, E. (2019). The effects of nontariff measures on Agri-food trade: A review and metaanalysis of empirical evidence. Journal of Agricultural Economics, 70(3), 595-617. https://doi.org/10.1111/1477-9552.12316
  50. Schlueter, S. W., Wieck, C., & Heckelei, T. (2009). Regulatory policies in meat trade: Is there evidence for least trade-distorting sanitary regulations? American Journal of Agricultural Economics, 91(5), 1484-1490. http://www.jstor.org/stable/20616330. https://doi.org/10.1111/j.1467-8276.2009.01369.x
  51. Silva, J. M. C. S., & Tenreyro, S. (2006). The log of gravity. Review of Economics and Statistics, 88(4), 641-658. https://doi.org/10.1162/rest.88.4.641
  52. Sung, J. C., Saera, O., & Sang, H. L. (2020). The impacts of structural similarity of nontariff measures on agricultural trade. Sustainability, 12(24), 10545-10558. https://doi.org/10.3390/su122410545
  53. Tinbergen, J. (1962). Shaping the world economy: Suggestions for an international economic policy. http://hdl.handle.net/1765/16826
  54. Tran, N., Wilson, N., & Hite, D. (2013). Choosing the best model in the presence of zero trade: A fish product analysis. In J. C. Beghin (Ed.), Nontariff measures with market imperfections: Trade and welfare implications (pp. 12-35). Bingley: Emerald Group Publishing Limited. https://doi.org/10.1108/S1574-8715(2013)0000012010
  55. UNCTAD. (2017). TRAINS NTMs: The global database on nontariff measures. https://trains.unctad.org/Forms/Analysis.aspx
  56. UNCTAD. (2019). International classification of non-tariff measures. https://hdl.handle.net/20.500.12870/1593
  57. USDA Economic Research Service. (2021). Rice sector at a glance U.S. https://www.ers.usda.gov/topics/crops/rice/rice-sector-at-a-glance
  58. Valera, H., Yamano, T., Pede, V., & Balie, J. (2021). Potential trade policy reforms in Southeast Asian rice markets: Domestic and international impacts. https://ricetoday.irri.org/potential-trade-policy-reforms-in-southeast-asian-rice-markets-domestic-and-international-impacts/
  59. Webb, M., Gibson, J., & Strutt, A. (2019). Market access implications of non-tariff measures: Estimates for four developed country markets. World Economy, 42(2), 376-395. https://doi.org/10.1111/twec.12706
  60. Webb, M., Strutt, A., Gibson, J., & Walmsley, T. (2020). Modeling the impact of non-tariff measures on supply chains in ASEAN. World Economy, 43(8), 2172-2198. https://doi.org/10.1111/twec.12955
  61. Will, M. (2020). Making Gravity Great Again Policy Research Working Paper, No. 9391, World Bank, Washington, DC. http://hdl.handle.net/10986/34477
  62. World Integrated Trade Solution. (2021). Tariff and trade analysis. https://wits.worldbank.org/WITS/WITS/Restricted/Login.aspx
  63. World Trade Organization (WTO). (2012). A practical guide to trading policy analysis. https://www.wto.org/english/res_e/publications_e/practical_guide12_e.htm
  64. Xiong, B., & Beghin, J. (2012). Does European aflatoxin regulation hurt groundnut exporters from Africa? European Review of Agricultural Economics, 39(4), 589-609. https://doi.org/10.1093/erae/jbr062
  65. Xiong, B., & Beghin, J. (2017). Disentangling demand-enhancing and trade-cost effects of maximum residue regulations. World Scientific, 14, 105-108. https://doi.org/10.1142/9789813144415_0006