Ilham Febri Budiman, Nindyta Lutfia Annur, Rachel Jenny Monintja
This study evaluates the structural determinants of Indonesia's manufacturing exports to Regional Comprehensive Economic Partnership (RCEP) member countries from 2007 to 2024. By utilizing a machine learning approach through the Random Forest algorithm within the Knowledge Discovery in Databases framework, this research transcends the rigid linearity assumptions of traditional econometric models. The computational model demonstrates superior predictive precision, with the model’s accuracy reaching a score of 0.983. A specialized variable importance analysis reveals that economic similarity is the primary catalyst for export flows, thereby strongly confirming the Linder hypothesis regarding intra-industry trade. Furthermore, fundamental gravity model variables, namely economic distance, population size, and gross domestic product, remain highly significant in dictating bilateral trade volumes. Interestingly, the empirical results indicate that the implementation of the RCEP pact and the global pandemic shock possess negligible immediate impacts on the export volume. This anomaly suggests a substantial time lag effect, as domestic industrial supply chains require a considerably long adaptation period to optimally capitalize on tariff eliminations. Consequently, this paper recommends strategic policy interventions for fiscal and customs authorities. The government must synergize logistical infrastructure improvements with targeted industrial incentives, particularly the super tax deduction for research and development and the Import Facility for Export Purposes. These synchronized efforts are essential to reduce economic distance friction, enhance structural competitiveness, and transform Indonesia's participation in the RCEP agreement into tangible manufacturing trade creation.
Article Details
| Volume: | 6 |
| Issue: | 2 |
| Year: | 2026 |
| Published: | 2026-06-28 |
| Pages: | 311–327 |
| Section: | Articles |

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
This work is licensed under a Creative Commons License.
Track citations and research impact