A Multifaceted Panel Data Gravity Model Analysis of Peru’s Foreign Trade

Xu Wang, Ryan P. BadmanKSP 2016

TL;DR

This paper applies an augmented gravity model to analyze Peru's export performance, revealing that economic size and trade agreements are key drivers, while distance acts as a significant barrier. The study provides specific insights into Peru's copper exports and the impact of regional trade blocs like MERCOSUR, offering policy recommendations for sustainable economic growth.

What Problem Does This Paper Address?

Peru's economy has seen significant growth driven by its expanding export sector. However, understanding the precise factors influencing this export performance is crucial for developing effective economic policies. While general trade models exist, they might overlook specific nuances of Peru's trade landscape, such as the dominance of certain export commodities like copper or the impact of regional trade agreements. This paper aims to fill this gap by conducting a multifaceted analysis of Peru's export performance using an augmented gravity model, breaking it down into three key areas: general economic factors, the specific case of copper exports, and the influence of regional trade agreements.

Methods

The study employs an augmented gravity model of trade, a standard econometric tool for analyzing bilateral trade flows. The model is extended to include various factors beyond just economic size and distance, such as common language, common borders, and participation in regional trade agreements (APEC, CAN, MERCOSUR). The researchers utilize panel data analysis, applying fixed effects, random effects, and instrumental variables Generalized Method of Moments (IV-GMM) regression techniques. The choice of IV-GMM is particularly important as it addresses potential endogeneity issues and provides accurate estimates in the presence of heteroskedasticity. For the specific analysis of copper exports, the model is further modified to account for the unique characteristics of natural resource trade and the influence of major importing countries.

Datasets

The study utilizes annual panel data for Peru and its trading partners.

  • General Model of Peru (GMP): Data covers the period 2006-2015, including 108 countries.
  • Copper Trade of Peru (CTP): Data covers the period 2006-2015, focusing on 21 countries.
  • Regional Trade Agreement Export Analysis (RTP): Data covers the period 1994-2015, specifically focusing on MERCOSUR member and associate member countries.

Data sources include the IMF Direction of Trade Statistics, UN Commodity Trade Statistics (UN Comtrade), World Bank's World Development Indicators, and data from CEPII.fr for distance and common language.

Key Insights and Ideas

  • Economic Size Matters, Distance Hinders: Consistent with gravity model theory, larger economies (higher GDP) for both Peru and its trading partners positively influence export volumes. Conversely, physical distance significantly increases transportation costs and hinders trade.
  • Copper's Dominance and Unique Dynamics: Peru's copper exports exhibit distinct patterns. While Peru's GDP is a key determinant, the influence of specific importing countries (like Korea, China, and Japan) is so strong that it can overshadow general economic factors. This highlights the need for tailored analysis for resource-based exports.
  • Regional Trade Agreements Boost Trade (When Isolated): While the general model showed mixed results for MERCOSUR, a focused analysis on MERCOSUR members revealed a positive and significant impact on Peru's exports. This suggests that regional blocs can indeed foster trade, but their effect might be masked in broader analyses dominated by larger non-member trading partners.
  • Linder Hypothesis Supported: The negative and significant coefficient for the GDP per capita difference supports the Linder Hypothesis, suggesting that countries with similar income levels tend to trade more with each other, likely due to similar production structures and consumer preferences.

Key Results

  • General Exports (GMP): Peru's GDP, importer's GDP, and participation in APEC had positive and statistically significant effects on exports. Distance had a significant negative impact. The effect of common language and border was found to be insignificant in the general model, likely due to the dominance of non-Spanish speaking trading partners.
  • Copper Exports (CTP): Peru's GDP positively influenced copper exports. Inflation in importing countries (with a lag) was also positively significant. The common border had a negative and significant effect, which is counterintuitive and attributed to the unique nature of natural resource trade and potential political considerations. Specific country dummies revealed that Korea, China, and Japan were significant importers of Peruvian copper, while the USA imported significantly less.
  • Regional Trade Agreements (RTP): When analyzing only MERCOSUR and CAN members, participation in these blocs was found to have a positive and statistically significant impact on Peru's exports. This corrected the bias observed in the general model.

Limitations and Future Work

  • Data Limitations: The study notes that for some analyses, particularly the RTP, data availability for certain regional organizations (like CAN) was limited, leading to a focus on MERCOSUR.
  • Masked Effects: The general model's results for regional trade agreements were potentially masked by the dominance of non-regional trading partners. This highlights the importance of disaggregated analysis.
  • Natural Resource Specificity: The copper export analysis suggests that for natural resources with a few dominant trading partners, general economic mass might not be a sufficient explanatory variable, necessitating the inclusion of country-specific demand factors.
  • Future Work: The authors suggest that further research could explore the impact of long-term contracts and cultural tendencies on trade flows. They also recommend continued active participation in MERCOSUR and APEC, and investment in transport and communication infrastructure to mitigate the negative effects of distance. For the copper sector, developing production and fostering long-term cooperation with major importers like China, Korea, and Japan is advised to manage dependence.

Summary updated

Abstract

Peru’s abundant natural resources and friendly trade policies have made the country a major economic player in both South America and the global community. Consequently, exports are playing an increasingly important role in Peru’s national economy. Indeed, growing from 13.1% as of 1994, exports now contribute approximately 21% of the GDP of Peru as of 2015.Given Peru’s growing global influence, the time is ripe for a thorough analysis of the most important factors governing its export performance. Thus, within the framework of the augmented gravity model of trade, this paper examines Peru’s export performance and attempts to identify the dominant economic factors that should be further developed to increase the value of exports. The analysis was conducted from three different aspects: (1) general economic parameters’ effect on Peru’s export value, (2) more specific analysis into a major specific trade good, copper, and (3) the impact that regional trade agreements have had on Peru’s export performance. Our panel data analysis results for each dataset revealed interesting economic trends and were consistent with the theoretical expectations of the gravity model: namely positive coefficients for economic size and negative coefficients for distance. This report’s results can be a reference for the proper direction of Peruvian economic policy so as to enhance economic growth in a sustainable direction. Keywords. Gravity model, Fixed effect, Random effect, Instrumental variables GMM regression, Exports, Copper, Panel data, Peru, Regional trade agreements. JEL. F10, F12, F40, F41.
BibTeX
@inproceedings{wang2016multifaceted,
  author = {Xu Wang and Ryan P. Badman},
  title = {{A Multifaceted Panel Data Gravity Model Analysis of Peru’s Foreign Trade}},
  booktitle = {KSP},
  year = {2016},
}
Figure