Do food safety regulations impede agrifood exports of China?

1 INTRODUCTION

Import tariff has significantly declined after multiple rounds of trade negotiations at the World Trade Organization (WTO) and the enactment of various trade agreements by trade partners (Aksoy & Beghin, 2004). Resultantly, many developing countries are participating in the global trading system. However, tariff duties are being replaced by the developed countries through non-tariff barriers (NTBs), for example, technical barriers to trade (TBTs) and sanitary and phyto-sanitary (SPS) measures to restrict imports from developing countries especially food and agricultural products. These measures could cut or counterbalance the returns of tariff reduction (WTO, 2022).

Because of increased global trade, the food safety issue is now considered a global issue. Maximum residue limits/levels (MRLs) is one of the measures implemented by developed countries to check imports, especially food and agricultural commodities for their quality. The MRLs are implemented by importing countries to safeguard the health of consumers, the environment, and most important domestic industry. However, these measures mostly restrict exports of agriculture and food products from developing to developed economies (Grundke & Moser, 2019; Jongwanich, 2009; Kinzius et al., 2019; Li & Beghin, 2012, 2017; Neeliah et al., 2013). NTMs also dampen the effect of preferential treatment agreements (PTAs) and reduce the extensive margins of trade (entry to markets), indicating that in the presence of NTMs, tariff preferences are insufficient to access international markets (Cipollina & Demaria, 2020). Bao and Qiu (2010) and Moenius (2006) report that NTBs cut agricultural exports and promote manufacturing goods. While Ishaq et al. (2016) infer that food safety regulations (MRLs) have a trade-promoting impact on exports of food commodities. Therefore, Bao and Qiu (2012) conclude that NTMs may have both trade promotion and demotion effects depending on their settings. Moenius (2006) finds that meeting these standards increases exporting costs it also facilitates and lowers search costs for both consumers and producers. Disdier et al. (2008) consider the vague effects of SPS and TBTs on trade. Because they find zero effect for trade between OECD (developed) countries and negative effects for exports from developing to OECD countries. While explaining the effects of stringent food safety regulations from a consumer point of view Alazzam (2021), Disdier et al. (2008), and Ishaq et al. (2016) conclude trade promoting effects if consumers are satisfied with the product. Researchers like Drogué and DeMaria (2012), Henson and Jaffee (2008), Jongwanich (2009), Li and Beghin (2012), and Li and Beghin (2017) conclude their findings with trade impeding impacts of NTBs, especially on exports from developing countries.

Santeramo and Lamonaca (2019) conducted a meta-analysis of NTMs’ theoretical and empirical studies to conclude the impact of NTMs on trade. They analyzed 62 papers encompassing 1362 observations and 1213 estimated t-statistics. The study used a combination of a multinomial logit model, a Probit model, and a Tobit model to estimate important factors affecting the direction of trade impacts, the probability of an impact coefficient being significant, and the accuracy of significant estimates. Most studies (56%) produced primarily negative impacts (34%) found both negative and positive impacts, while only 6% found a positive trade effect, and 4% found no significant effects. Moreover, the magnitude, direction, statistical significance, and accuracy of trade impacts were case-specific, greater for types of NTMs but lower for proxies of NTMs compared to the average effect. MRLs and ad valorem equivalent enhance trade, SPSs have mixed results, and TBTs have trade-limiting effects.

Despite its major role in the international trading system1 especially after accession to the WTO, the food safety issue is frequently occurring in China. Ishaq et al. (2014) describe the frequent occurrence of food safety issues in China has shaken the trust of consumers and tagged “Made in China” with a bad name both in local and international markets. The event of melamine contamination in baby milk formula jolted the whole country. Food safety issue in China has received much attention from researchers both at home and abroad. Appendix A depicts a brief review of some of the recent studies covering the NTBs. This figures out that among others, Li and Beghin (2017), Ishaq et al. (2016), Sun et al. (2014), Ishaq et al. (2014), Song and Chen (2010), Bao and Qiu (2010), Ni and Zeng (2009), Huang et al. (2010), Chen et al. (2008), are the efforts highlighting the issue of Chinese food trade. After the melamine incident (October 2008) the government revisited its food safety policy and adopted strict measures to curb food safety issues in future.

This study aims to estimate the effects of food safety regulations (the stringency of MRLs) levied by the partner countries on major exports2 of agrifood products from China. However, this study differs from previous works in number of ways. First, this study is taking into account the current MRLs adopted by the government of China to safeguard consumers. Second, this study covers major food exports of the country. Third, rather than concentrating on one or two substances/pesticides, this study covers the whole range of MRLs for a particular commodity. Fourth, this study employs indices developed by Li and Beghin (2014) for calculating MRLs to empirically show their impact on agrifood exports of China.

The panel data are estimated using both the fixed- and random-effect models. Results of the study show that all the standard gravity-type variables carry signs according to prior expectations and economic theory. The study concludes that food safety regulations have no effect on exports of all the selected commodities except meat of swine.

2 METHODOLOGY

2.1 The model

This study estimates the gravity trade model to highlight the impact of food safety regulations (MRLs) regulated by the partner countries to check the exports of major agrifood products from China. The gravity trade model is derived from Newton's Law of Gravitation. According to this model, the trade volume between two trading partners is directly proportional to its economic sizes (GDP/per capita income) and inversely proportional to the distance between these two countries. Moenius (2006) took the lead and report the impact of standards on bilateral trade. After his work, Sun et al. (2014), Wilson and Otsuki (2003) also estimated the gravity equation to address this issue. Among others Chen et al. (2008), Disdier and Marette (2010), Drogué and DeMaria (2012), Ishaq et al. (2016), Melo et al. (2014), Sun et al. (2014), Wei et al. (2012), Xiong and Beghin (2012) estimate the gravity trade model to highlight the effects of food safety regulations on bilateral trade flows.

The gravity model is:

()

where is the export value of commodity from country to while , , and represent exporting country, importing country, and resistance to trade flows, respectively.

The gravity model in the log-linear version is:

()

where is the value of export (measured at 1000 $US) of commodity from exporting country-China to the trading partner at time , is the production (measured at 1000 tonnes) of commodity in exporting country-China at time , is the value of per capita GDP (measured at 10,000 $ US) of country at time , represents the stringency of food safety regulation and is the index of Maximum Residue Limits levied by the trading partner on exports of commodity from China at time , is the distance (measured at 1000 km) between the capital cities of exporting country-China and the importing country . However, are dummies and equal to 1 if country is landlocked, if China and country share a common language, and if China and the importing country share a common border, respectively. While are respectively the fixed effects for country , year and commodity . are the coefficients to be estimated and is the error term.

is estimated using the following formula3:

()

where is the index of maximum residue limits on commodity , is the total number of chemicals applied on commodity , and are respectively the MRLs determined by the Codex Alimentarius and importing country for commodity and substance . If it represents the non-protectionist policy, if represents the protectionist policy, and if represents anti-protectionism policy. When the MRLs stringency for a particular commodity in a respective country is similar to that of Codex Alimentarius, trade policy is regarded as nonprotectionist. Wherein, protectionist trade policy means that MRLs for a particular commodity in a particular country are more rigid than that of Codex Alimentarius and anti-protectionism policy means that MRLs for a particular commodity in a particular country are less rigid than that of Codex Alimentarius.

3 RESULTS AND DISCUSSION

The gravity model (Equation 2) is estimated using both random and fixed effects techniques. FE models assume that something within the entity may affect or bias the trade volume. These features are exclusive to the entity and should be uncorrelated with other entity. FE models get rid of the time-invariant characteristics to analyze the net effect of trade volume. These models have high standard errors assuming that the error term is correlated with predictor variables. FE models are not effective if there is low variation within cluster or variables vary gradually over time or when the error term is correlated with the constant (Hausman test). In this case, the random-effects (RE) model should be used. RE models assume that unobserved and observed variables are not correlated. This assumption may not hold often. RE models have low standard errors as compared to FE.

Separate models are estimated for each commodity using both the FE (Table 1) and RE (Appendix B) techniques. Results show that the estimated models fit the data well, the coefficient of variation (R²) value ranges from 91% for exports of kidney beans to 95.5% for export of mandarins using the FE technique. The R² value for estimated models is low in the case of the RE technique ranging from 13.6% for kidney beans to 55.1% for meat of swine. The probability of the Hausman test is 0.0354 which is significant at a 0.05 level and supports the estimation of the FE model.

The model specified in Equation (2) contains log of MRL value to capture the effect of stringent regulations on major export commodities. FE models show that MRLs have no effect on exports of all commodities except meat of swine which is negatively affected by the stringent regulations of importing countries. Similarly, RE models also show the same results except for garlic and mandarin. Wherein with the imposition of MRLs exports of both commodities have increased. Our results are differing the findings of previous researchers, for example, Chen et al. (2008), Jongwanich (2009), Bao and Qiu (2012), Drogué and DeMaria (2012), Li and Beghin (2012), Melo et al. (2014). They infer that food safety standards have a trade-distorting impact on the export of agrifood commodities from developing countries. Ishaq et al. (2016) conclude that MRLs have a trade-promoting impact on exports of food commodities. In other words, differences in regulations between developed and developing countries matter and hinder trade flow. Furthermore, the results of this study are in line with the findings of Schuster and Maertens (2015) and Xiong and Beghin (2012) who empirically show that private standards have no effect on export performances of firms, for example, export volumes, export values, at the intensive margin, and at the extensive margin. The results are not wondering as China has adopted a strict policy to ensure food safety; country's current level of MRLs is more stringent than many of the developed countries (Appendix C). Because in China, the Ministry of Health has developed more than 4800 food standards and integrated more than 1200 standards in consultation with line ministries and relevant industries. These integrated standards include contaminants, dairy products, food additives, food products, general basic standards, hygiene standards, mycotoxins, nutritional supplements, pesticides, prepackaged food labels and nutrition labels, test methods, wine, and other special standards. Further, these organizations produced about 12,000 key indicators of raw materials to produce safe processed food, and protect the environment and plant and animal health Liu et al. (2019).

The estimated models also contain per capita GDP in importing countries, distance and dummy variables for common borders, common language, landlocked, colonial relationships (Colony), and preferential trade agreements (PTAs). Per capita GDP indicates the income level and has a negative sign for meat of swine and mandarins. This indicates that exports of these products decrease as the income level in the importing countries increases. Distance is used as a proxy to represent transportation costs. According to theory, distance is anticipated to have a negative sign because an increase in distance between the trading countries means higher transportation costs and resultantly lesser trade volume. The effect of distance for the estimated models is negative and statistically significant in FE models for all commodities except garlic while nonsignificant in RE models except garlic and apples. Similarly, trading partners with a common border and common language are supposed to trade more. Border and common language variables have positive signs with different significance levels. In the same pattern, it is expected that countries having colonial relationships and signatories of trade agreements, trade more. The colony variable is positive and significant only for meat of swine and tea exports in FE models. While PTA has a significantly negative sign for meat of swine in case of FE models. This may be due to the fact that the volume of trade is more to those countries which are not signatories of trade agreements with China. Generally, it is believed that transportation through seawater is cheaper and countries have access to seawater trade more. In other words, trade flow is low if one or both of the trading partners are landlocked. In this case under all the models, the landlocked variable has a negative sign and mostly significant. It implies that seawater has an important role in shaping exports.

4 CONCLUSION

This study aims to estimate the effect of maximum residue limits levied by the trading partners on the exports of agrifood commodities from China. The gravity model is estimated for products for which MRLs data are available namely apples, garlic, kidney beans, mandarins, meat of swine, and tea. The study covers annual bilateral trade data from 2009 to 2013. Results show that all the standard gravity-type variables carry signs according to prior expectations. Furthermore, FE models show that MRLs have no effect on exports of all commodities except meat of swine which is negatively affected by the stringent regulations of importing countries. Similarly, RE models also show the same results except garlic and mandarins.

AUTHOR CONTRIBUTIONS

Muhammad Ishaq: Conceptualization; formal analysis; methodology; writing—original draft; writing—review and editing. Zahoor ul Haq: Data curation; methodology; writing—review and editing. Ping Qing: Supervision; validation. Chongguang Li: Data curation; supervision; validation.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

ETHICS STATEMENT

This research is based on data extracted from different sources and all the sources are cited with the data of accession if needed. This work also acknowledges the works of other authors used in any part of this manuscript in accordance with the standard format of the journal.