Swine meat is a crucial component of global food security, comprising over one-third of world meat production. However, risks exist, notably from unpredictable animal infectious diseases like African swine fever (ASF). The 2019 ASF outbreak in China, resulting in the deaths of 180 million pigs, dramatically illustrates the market instability these events can cause. Food autarky, or self-sufficiency, is a high-priority policy for many food-deficit nations, particularly after the 2008 food price crisis. While the effectiveness of this policy is debated, there is a lack of comprehensive quantitative analysis, especially regarding its impact on cross-border price transmission. This study addresses this gap by examining the causal relationships and dynamic conditional correlations (DCCs) between global and local pork prices in ten net importing countries using GARCH-type models and panel regression models. The goal is to identify factors, such as self-sufficiency rates and trade volume, affecting the price volatility transmission and the effectiveness of self-sufficiency policies in insulating domestic markets.

Existing literature on food autarky policy often lacks quantitative analysis. Some studies have used stochastic computable general equilibrium models and GARCH-type models with DCC to examine the impact of self-sufficiency on price volatility transmissions in cereal and beef markets. These studies generally suggest that higher self-sufficiency insulates domestic markets from foreign shocks, but Tanaka and Guo (2019) suggest domestic prices may become more unstable if domestic supply is more volatile than foreign supply. Regarding price transmission, much research focuses on domestic price linkages, while relatively few studies investigate international market connections, particularly in the meat sector. A knowledge gap exists regarding the determinants of cross-border food price transmissions, especially the role of food autarky policies in the pork market.

Monthly data from January 2001 to December 2018 were collected for international and local pork prices in ten net pork-importing countries: China, Colombia, Japan, Mexico, New Zealand, Peru, Philippines, Romania, South Korea, and the United Kingdom. The data underwent seasonal adjustment using the X-13-ARIMA method. Unit root tests confirmed the stationarity of the price return series. Univariate GARCH-type models (GARCH, GJR-GARCH, and EGARCH) were employed to estimate the volatility of each price series, selecting the best model using the Schwarz Information Criterion (SIC). Granger causality tests were performed using Hong's (2001) method to assess causality-in-mean and causality-in-variance. A bivariate GARCH model with DCC was used to estimate the time-varying correlations between international and domestic pork prices. The DCC models included standard DCC, A-DCC, G-DCC and AG-DCC models. Finally, panel regression models, including pooled OLS, fixed effects, random effects, FGLS, and Prais-Winsten with PCSEs, were used to investigate the effects of self-sufficiency rates (SSR), trade volume rates (TVR), and beef consumption on DCCs, with yearly data for SSR and TVR. Country-specific analysis was conducted using a random-coefficient linear regression model to check for heterogeneity.

Granger causality tests revealed diverse relationships between global and local pork prices across countries. China exhibited a unidirectional causality-in-mean effect from international to domestic prices, suggesting rapid transmission. Conversely, Colombia, South Korea, Mexico, Peru, Philippines, Romania, and the United Kingdom showed a causality-in-mean effect from domestic to international prices. Regarding causality-in-variance, China and Colombia showed unidirectional causality from international to domestic markets, while Romania demonstrated bidirectional causality. The United Kingdom showed a unidirectional causality-in-variance from domestic to international markets. Time-varying conditional correlation (DCC) analysis revealed varying patterns across countries, with some showing strong co-movement during periods like the 2007-2009 global financial crisis. The mean and median DCC values were generally low, but South Korea showed the highest and the United Kingdom the lowest. Panel regression analysis indicated that higher self-sufficiency rates (SSR) were significantly and negatively associated with DCCs. Higher trade volume rates (TVR) were positively associated with DCCs. Beef consumption had a significant negative effect on DCCs across various models, suggesting a substitutive effect. Country-specific analysis using random-coefficient regression generally confirmed these findings, although exceptions existed for China and the United Kingdom.

The findings suggest that policies aimed at increasing self-sufficiency in pork can mitigate the impact of international price volatility on domestic markets. This is particularly relevant for food security, especially for low-income households. However, the study highlights that increasing self-sufficiency doesn't automatically lead to better market outcomes; it is crucial to weigh the efficiency gains from open markets against the stability achieved by higher self-sufficiency. The positive relationship between TVR and DCC underscores the importance of international trade in price transmission. The negative impact of beef consumption on DCC suggests that diversifying food consumption can act as a buffer against external price shocks. The country-specific analysis revealed some exceptions, illustrating the heterogeneity in market dynamics. The different responses of China and the United Kingdom could be due to their significant market shares and roles in influencing global prices.

This study contributes to the understanding of cross-border price spillovers in the pork sector. The findings highlight the mitigating role of self-sufficiency and the buffering effect of substitute goods like beef. Further research could explore the cost-benefit trade-off of self-sufficiency policies, incorporating risk aversion measures to translate volatility reduction into monetary values, and examining the dynamics of internal price volatility within each country.

The study did not consider internal price volatility (e.g., wholesale or producer prices), limiting the complete assessment of self-sufficiency's impact. The findings might not directly apply to exporting countries, where high self-sufficiency could lead to higher trade volumes. The selection of countries for the panel data analysis may not be fully representative of the broader international pork market.