The impact of mutual recognition of geographical indications on the quality upgrading of China's agricultural exports

China is the world's largest producer and trader of agricultural products, yet its agricultural exports are "big but not strong" and remain in a transition from quantity- to quality-driven growth, facing a dilemma of quality upgrading. Geographical Indications (GIs) are a form of intellectual property identifying origin and quality attributes and can create regional public brands that enhance export quality and prices. Since 2007, China and the EU initiated a "10+10" pilot of mutual GI recognition, with cross-certification completed in 2012. The study asks whether mutual GI recognition improves the quality of China’s agricultural exports, through which supply or demand-side mechanisms, and whether effects vary across markets, products, and firms. The paper builds a theoretical framework in which mutual GI recognition differentiates products, potentially expanding markets and changing incentives on both supply and demand sides. It then empirically tests the impacts using a multi-time point DID model on Chinese customs data (2000–2016), decomposing mechanisms and examining heterogeneity. Contributions include shifting the perspective from unilateral GI certification to bilateral mutual recognition, analyzing dual supply- and demand-side channels, and enriching theories of quality upgrading and comparative advantage in agricultural trade.

Two strands of literature are reviewed. First, on the quality of exported agricultural products: Early theories (Linder, 1961; Flam and Helpman, 1987) linked quality to income and vertical differentiation; heterogeneous firm trade models (Melitz, 2003) opened analysis of quality variability. Empirical measures include unit values (Schott, 2004), backcasting/post hoc methods (Khandelwal et al., 2013; Shi and Shao, 2014), and nested logit approaches (Dong and Huang, 2016). Influences on quality include SPS measures and MRLs that can hinder quality upgrading (Dong and Liu, 2019; Chen and Xu, 2017; Jiang and Yao, 2019) versus factors like FDI, better institutions, and reduced policy uncertainty that support quality and value upgrading (Ciani and Imbruno, 2017; Hu and Zhao, 2018; Sun and Anwar, 2019). Collective reputations can lead to free-riding by low-quality firms (Cage and Rouzet, 2015), though regional reputation can also foster upgrading via competition and spillovers (Dong and Gao, 2020). Second, on GIs and trade: High-quality GI producers export more and at higher prices (Crozet et al., 2012); GI certification facilitates exports within and beyond the EU, though effects may weaken where importers have many GIs (Raimondi et al., 2020; Filippis et al., 2022). Some argue GI rules can raise costs and reduce competitiveness (Moschini et al., 2008; Bienabe and Marie-Vivien, 2017). GI certification can spur innovation and quality improvements (Merel and Sexton, 2012; Agostino and Trivieri, 2014). Gaps include limited study of bilateral GI mutual recognition (as opposed to unilateral certification), scant evidence on developing-country firms (especially China), and under-explored demand-side mechanisms in destination markets.

The study develops a theoretical model embedding mutual GI recognition into a heterogeneous-firm framework (building on Melitz, 2003; Antoniades, 2015), defining demand for differentiated products with quality preferences and supply with technology-driven quality improvement and cost changes. Hypotheses: H1 mutual GI recognition improves export product quality; H2 specialization agglomeration channel; H3 cost-saving channel; H4 domestic demand upgrading channel; H5 product recognition channel. Empirically, a multi-time point difference-in-differences (DID) model is estimated using Chinese customs firm–product–destination export data to EU member states from 2000–2016. The specification is: Y_jt = α + β (Treat × Post)_jt + Controls + firm–year, product–year, and destination–year fixed effects + error. Parallel trends are tested via event-time interactions. To address potential selection, a PSM-DID approach matches on SPS measures, openness, per capita GDP, and GI endowment differences, using nearest-neighbor matching year-by-year. Robustness checks include alternative dependent variable (dummy for quality increase), trimming/truncation of quality outliers, alternative matching methods (caliper and kernel), and a placebo test with random assignment of treated products. Quality measurement follows Khandelwal et al. (2013)/Shi (2014): regress log quantity on log price with product and time effects to recover residual quality; transform residuals into a normalized quality index and compute period-to-period differences (Δquality) as the outcome. Policy/treatment: Treaty_jt indicates whether product is in China–EU mutual GI recognition list; Post_fjkt indicates treatment period by product–destination–firm–year; Treat × Post is the interaction. Controls: importing-country SPS notifications (lag HS2-level), openness (trade/GDP), per capita GDP, RMB exchange rate, and GI endowment differences by HS codes. Data: Firm–product (HS6)–destination export data from China Customs; SPS from WTO SPS notification system; macro controls from World Bank WDI; GI data for China from CNIPA, SAMR, MOA, and Geographical Indication Network; EU GI data from eAmbrosia. Sample includes EU-28 over 2000–2016, covering major agricultural HS chapters (e.g., HS02, HS03, HS04, HS07, HS08, HS09, HS19, HS22), totaling 240,814 observations (baseline regressions N ≈ 133,785 after fixed-effects sample constraints). Endogeneity is further addressed via 2SLS using instruments at HS2 level: average number of GIs and agricultural export value, argued relevant for pilot selection, with first-stage relevance and over-identification/weak-IV diagnostics reported. Mechanism tests use mediators: specialization/agglomeration (number of new products exported; share of processed products—product restructuring), cost saving (number of EU export destinations; fixed trade costs proxy), domestic demand upgrading (domestic demand quality via divergence index between domestic consumption and exports), and product recognition (import demand elasticity estimated from double-log demand with destination GDP per capita control). Two-step mediator models estimate effects of Treat × Post on mediators and mediators on Δquality with the same fixed effects.

- Mutual GI recognition significantly increases the quality of China’s agricultural exports to the EU. In baseline PSM-DID regressions with two-way fixed effects, Treat × Post is positive and statistically significant (e.g., coefficient positive and significant with t-statistics around 3.4–4.4). Event-study indicates no pre-trends and rising post-treatment effects. Placebo tests with random treatment assignment yield insignificant effects, supporting identification. IV-2SLS using HS2 GI endowment and export value as instruments confirms a positive causal impact; first-stage F statistics are large (e.g., KP-LM > 600; Wald F > 100), mitigating weak-IV concerns. - Robustness: Results hold when (i) replacing Δquality with a quality-improvement dummy, (ii) trimming/truncating outliers, and (iii) using alternative PSM matching methods (caliper, kernel). - Heterogeneity: (i) Importer GI endowment—effects are stronger in high GI-endowment EU countries (UK, France, Spain, Italy), consistent with greater consumer familiarity and willingness to pay. (ii) Firm scale and proximity to the quality frontier—effects are larger for medium/large firms (significant at 50% and 90% quantiles) and for firms closer to the world quality frontier; firms far from the frontier benefit less. (iii) Product factor intensity—significant effects for labor-intensive categories where China has comparative advantages (vegetables HS07, fruits HS08, beverages HS22); no significant effect for resource-intensive categories (e.g., meat HS02, cereals HS19) constrained by land/water resources and technology. - Mechanisms supported: (i) Specialization/agglomeration—mutual recognition reduces the range of new products (negative effect on count of new products) and increases the share of processed products; both channels correlate with higher export quality, indicating product concentration and upgrading. (ii) Cost saving—mutual recognition increases the number of EU destinations served and reduces proxies for fixed trade costs; more destinations and lower costs are associated with higher export quality. (iii) Domestic demand upgrading—mutual recognition raises domestic demand quality (divergence index), which in turn promotes export quality. (iv) Product recognition—mutual recognition lowers import demand price elasticity (consumers accept higher prices), and lower elasticity is associated with higher export quality. - Data coverage: 240,814 firm–product–destination–year observations (EU-28, 2000–2016); baseline regression samples around 133,785 observations after FE structure.

The findings directly address the core question: bilateral mutual recognition of GIs between China and the EU causally upgrades the quality of China’s agricultural exports. The theoretical channels—reduced trade costs, enhanced consumer quality preferences, and productivity/technology incentives—are borne out empirically. On the supply side, mutual recognition standardizes production, fosters specialized agglomerations, and lowers information and clearance costs, which reallocates resources toward quality-improving investments and concentrated product mixes. On the demand side, recognition enhances domestic demand quality and product reputation in destination markets, mitigating information asymmetries and enabling price premia, which incentivize firms to upgrade quality. The effects are context-dependent: strongest where GI awareness is high (EU high-endowment countries), among larger or more capable firms nearer the quality frontier, and in labor-intensive products where quality differentiation is feasible and less constrained by natural resources. These results enrich the literature on quality upgrading and comparative advantage by identifying bilateral GI mutual recognition as a policy instrument that leverages both supply- and demand-side forces to overcome quality traps.

The study develops a theoretical framework and provides robust empirical evidence that mutual recognition of geographical indications between China and the EU significantly improves the quality of China’s agricultural exports. It shows that the effects operate via supply-side specialization/agglomeration and cost-saving mechanisms, and demand-side domestic demand upgrading and product recognition, with stronger impacts in high GI-endowment markets, for larger and frontier-proximate firms, and for labor-intensive products. The research contributes to the understanding of how bilateral GI arrangements can alleviate quality upgrading dilemmas and enhance value-added trade in agriculture. Policy recommendations include: expanding GI mutual recognition with additional trading partners strategically; strengthening GI protection and oversight to shift from quantity to quality; and aligning domestic supply-side reforms with international quality benchmarks, standards, traceability, and digital quality management to support sustained upgrading of agricultural exports.