Data for The Andean-Amazonian Bioeconomy: A New Paradigm for Productivity and Well-being Beyond Traditional Economic Models

Description

The empirical analysis in this study is based on an unbalanced panel data set covering six Latin American countries over a 29-year period. This dataset was carefully constructed to capture the multiple dimensions of the Andean-Amazonian Bioeconomy. It includes key variables from three distinct sources: traditional socioeconomic indicators (such as per capita income and life expectancy), biophysical and environmental metrics (such as biocultural savings), and the central metric of our research, the Total Factor Productivity of the Andean-Amazonian Bioeconomy (TFP-BAA), derived from the application of the Malmquist Index. The combination of this data allows us to transcend conventional economic analysis and explore the interplay between economic progress, social welfare, and biophysical limits within the context of the "Vivir Bien" paradigm.

Steps to reproduce

Step 1: Data Preparation and Transformation • Variable Construction: Use the complement of the Gini coefficient (1 - G) to represent income equality. • Logarithmic Transformation: Apply natural logarithms to GDP per capital and the Social Welfare Index (HDI) to linearize the relationships and interpret results as elasticities. Step 2: Cross-Sectional Dependence Test • Action: Before testing for stationarity, apply the Pesaran CD Test. • Justification: If cross-sectional dependence exists, standard unit root tests may be biased. This justifies the later use of the Driscoll-Kraay estimator. Step 3: Unit Root Testing (TFPAAB / Fisher-ADF) • Action: Conduct the Fisher-type Augmented Dickey-Fuller (ADF) unit root test. • Procedure: 1. Test variables in levels: 𝑦𝑡 2. Test variables in first differences: ∆𝑦𝑡. • Goal: Confirm that all variables are integrated of order one, I(1), which is a prerequisite for cointegration. Step 4: Cointegration Testing (Westerlund / Pedroni) • Action: Apply the Westerlund (2007) or Pedroni tests for panel cointegration. • Goal: Reject the null hypothesis of "no cointegration" to prove a stable long term relationship exists between HDI, GDP, and the Gini complement. Step 5: Long-Term Equation Estimation • Model: Estimate the static long-term equation using Fixed Effects (FE). • Formula: l𝑛(𝐻𝐷𝐼𝑖𝑡) = 𝛽0 +𝛽1𝑙𝑛(𝐺𝐷𝑃𝑖𝑡) + 𝛽2𝑙𝑛(𝐸𝑞𝑢𝑎𝑙𝑖𝑡𝑦𝑖𝑡) + 𝜇𝑖 + 𝜀𝑖𝑡 • Validation: Extract the residuals to ensure they are stationary I(0). Step 6: Short-Term Dynamics and Error Correction Model (ECM) • Action: Estimate the Error Correction Model following the Engle-Granger two step approach. • Formula: Δ𝑙𝑛(𝐻𝐷𝐼}𝑖𝑡) = 𝛼 + ∑𝛾Δ 𝑋𝑖𝑡−1 +𝜆(𝐸𝐶𝑇𝑖𝑡−1) + 𝑒𝑖𝑡 • Key Parameter: The Error Correction Term (ECT) must be negative and statistically significant, representing the speed of adjustment (e.g., 9.64%). Step 7: Robust Inference (Driscoll-Kraay) • Action: Re-estimate the standard errors using the Driscoll-Kraay covariance matrix estimator. • Justification: This ensures that the t-statistics are robust against heteroskedasticity, autocorrelation, and cross-sectional spatial correlation.

Institutions

• Universidad Autonoma Chapingo

  Mexico, Texcoco
• Universidad Nacional Autonoma de Nicaragua Leon

  León, Leon
• Universidad Nacional Autonoma de Honduras

  Francisco Morazán, Tegucigalpa
• Universidad Mayor de San Andres

  La Paz, La Paz

Categories

Licence