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 11 / 23 PLOS One Trade agreement, climate change and geography The findings exhibit that exporting countries’ CO2 has a higher impact on bilateral export levels and probability than the importing country’s geography. Similarly, exporting countries’ physical distance has significantly impacted bilateral exports under outcome and selection equations. The export level and its probability are declined, resulting in the negative simultaneous impact of geography and the GDP. The marginal export level of exporting county’s CO2 has a positive impact by 247.6% and is greater than importing countries’ geography, conditionally varying the GDP level. Similarly, the marginal export probability of exporting country’s CO2 also has a positive impact of approximately 191.6% more than that of importing countries. Furthermore, the marginal export level and the marginal probability of physical distance are highly affected by approximately 264.09%, 204.56%, 14.494%, and 93.04% for both exporting and importing countries’ geography due to GDP. On conversely, [87] argue that trade is a benign effective factor for the environment in the short-run but harmful in the long-run. Moreover, CO2 emissions are negatively affected by urbanization in the short and long run. However, we find that the simultaneous impact of geographic factors has negative coefficients under outcome and selection equations. The complementary of CO2 and physical distance strongly impact the decline of the export level

and probability between trading countries. Table 3. Regression analysis. Exporting countries Importing countries ___-Estimation PPML OLS Fixed Effect Random Effect PPML OLS Fixed Effect Random Effect (1) (2) (3) (4) (1) (2) (3) (4) VARIABLES Dependent variable: exports CO2 0.102 * 0.020 -0.287 0.132 0.091*** 0.068 0.085** 0.084** (0.021) (0.054) (1.119) (0.477) (0.021) (0.058) (0.040) (0.040) Physical-distance 0.671** 0.859*** 0.414*** 0.525*** 0.541** 0.127*** 0.188*** 0.187*** (0.012) (0.016) (0.082) (0.072) (0.021) (0.018) (0.012) (0.012) Lndistance -0.762*** -0.840*** -0.857*** -0.858*** -0.671** -0.994*** -1.241*** -1.239*** (0.001) (0.030) (0.021) (0.029) (0.020) (0.031) (0.023) (0.024) Access to Ocean 0.201** 0.751*** – 0.622 0.210*** 3.006*** – 3.048*** (0.021) (0.044) – (0.386) (0.027) (0.039) – (0.404) Common border 0.021*** 2.416*** 2.609*** 2.606*** 0.032*** 2.248*** 1.501*** 1.506*** (0.091) (0.075) (0.082) (0.081) (0.032) (0.080) (0.066) (0.066) Lngdp 0.329** 0.281*** 0.429*** 0.445*** 0.450*** 0.644*** 0.582*** 0.582*** (0.017) (0.016) (0.041) (0.037) (0.030) (0.018) (0.013) (0.013) lntariff -0.039** -0.366*** -0.257*** -0.266*** -0.023*** -0.358*** -0.379*** -0.379*** (0.061) (0.022) (0.052) (0.051) (0.032) (0.023) (0.018) (0.017) lnpop 0.035*** 0.046*** 0.771*** 0.324*** 0.045*** 0.267*** 0.303*** 0.303*** (0.013) (0.016) (0.174) (0.103) (0.054) (0.015) (0.013) (0.011)

Constant 1.203** -1.818** -8.564 -8.226 1.4321** -5.726*** -3.407*** -5.266*** (0.021) (0.718) (13.700) (6.104) (0.042) (0.757) (0.542) (0.628) Observations 44436 44436 44436 44436 44344 44436 44436 44436 R-squared 0.405 0.104 – 0.349 0.426 – Number of exporters 46 – 46 46 46 – 46 46 Country effect Yes – – – Yes – – – Time effect Yes – – – Yes – – – Note: the term ***, ** and * denote that significance level at 1%, 5% and 10% respectively. Ln indicates logarithm for variables. https://doi.org/10.1371/journal.pone.0320363.t003