Climate change, geography and trade agreements: A perspective of Asian bilateral trade

PLOS One | https://doi.org/10.1371/journal.pone.0320363 April 25, 2025 9 / 23 PLOS One Trade agreement, climate change and geography contrast, several studies, e.g., [75–81] used standard gravity equations to analyze the influence of trade barriers in trade flows. Thus, this instrument is considered the most famous in foreign trade flows, such as imports and exports from origin to destination and vice versa, physical distance, economic mass (GDP), GNP, population size, culture, common language, and common border. To do so, econometric techniques have been employed, such as the Poisson Pseudo-Maximum Likeli- hood (PPML) to incorporate the issue of zero trade value in the dataset. The reason behind that some trading partners do not trade between each other such as Pakistan and Israel due to lack of diplomatic conflict or any economic and statically issue, which miss the observations [82–84] Furthermore, this approach is also used to tackle the issue of heterogeneity, which could be exist. Subsequently, fixed effect technique is used to estimate the constant or fixed relationship in the explanatory variables. Random effect analyses the data to evaluate certain factors which casually affect the outcomes across the individual or groups. Afterward, Iterative reweighted least squares and quantile regression are used to correct the robustness check. The reason behind that extreme values can create biasness in the analysis process. Empirical results Descriptive statistics. The descriptive statistics is analzyed in the Table 2 for measuring

the central tendency, rationality and effective technique. Analysis shows that the observations are almost closed to their means for each variable, suggesting that the variation is found moderate. For instance, the logarithm means and S.D values of the variable export are 6.045 and 6.774 respectively, subsequently, the logarithm means and S.D values of distance are 3.612 and 3.366, which indicate that lowest variation is existed. However, the CO2, Pop, and GDP variables have difference between mean and S.D values, indicating that there is some variation in these observations. Besides, there are much differences between maximum and minimum values for the variable’s exports, CO2, distance, population and GDP over time. Empirical analysis. We use different econometric techniques to incorporate the influence of geographic factors on bilateral exports. Table 3 reveals the estimates of PPML, OLS, FE, and RE on domestic and partner geographic factors in bilateral exports. The findings indicate that geographic factors have a positive impact on bilateral exports. The domestic log of CO2 Table 1. Variables and data source. Variable Measure Code Source Exports Ratio of export value to export price index EXP UNC- TAD Distance Kilometre from capital to capital Dist CEPII Tariff Weighted mean for all product in term of percentage TRF WDI Access to Ocean access to Indian or Pacific Ocean that defined as the equal to 1 if a country has access to Indian or the Pacific Ocean otherwise zero. AC CEPII

Common border Dummy variable, 1, if a country shared border with its trading partner, otherwise zero CB CEPII Carbon dioxide emissions Measured in tons released from all sectors CO2 WIOD Population Total population measured in million Pop WDI Gross domestic product Gross domestic product measured in billion US dollar GDP WDI Preferential trade agreement if exporter and importer listed were in any one of PTA’s listed either the WTO website PTA WTO WTOX four-dimensional (dummy) variable with 0 or 1 for each WTOX provision. WTOX WTO Source: author’s calculations. https://doi.org/10.1371/journal.pone.0320363.t001