Probabilistic and Asymmetric Effects of Green Accounting on Economic Prosperity and Economic Downturns Using Bayesian Logistic Regression

Technical Summary: Probabilistic and Asymmetric Effects of Green Accounting on Economic Prosperity and Economic Downturns

Problem Statement
This study addresses the gap in understanding the asymmetric effects of green accounting (GA) on different phases of the economic cycle, specifically distinguishing between periods of economic prosperity and economic downturns. While existing literature often treats green accounting as a monolithic factor or relies on single indicators (e.g., CO2 emissions, environmental taxes), this research argues that GA is a fragmented, multidimensional construct. Furthermore, prior studies have not sufficiently explored how the excessive expansion or severe contraction of green accounting implementation influences economic stability. The authors posit that while GA is generally intended to support sustainable development, its impact may vary probabilistically depending on the intensity of implementation and the prevailing economic context.

Methodology
The research employs a Bayesian Logistic Regression model to analyze panel data from 82 countries spanning the period 2002 to 2020.

• Variable Construction:

  • Green Accounting (GA): Instead of using single proxies, the authors constructed a composite Green Accounting Index using Principal Component Analysis (PCA). This index aggregates 16 component variables related to energy resources, green expenditures, internalization of environmental damage costs, and transparency in environmental taxation. The PCA reduced these variables into five principal components (retained based on eigenvalues > 1 and cumulative explained variance > 70%), which were then weighted by their explained variance ratios to form the final GA score.
  • Dependent Variables: Two binary variables were defined:
    • Economic Prosperity (EP): Equals 1 if the economic growth rate exceeds the sample average; 0 otherwise.
    • Economic Downturn (ED): Equals 1 if the economic growth rate is negative; 0 otherwise.
  • Asymmetry Analysis: To test for non-linear effects, the study introduced dummy variables for "excessive expansion" (GAH1, where GA > mean + 1.5 SD) and "severe contraction" (GAH2, where GA < mean - 1.5 SD), creating interaction terms to assess their specific impacts.
  • Control Variables: The model includes urbanization rate, unemployment rate, trade openness, inflation rate, and foreign direct investment.

• Modeling Approach:
  The study utilizes Bayesian inference to overcome limitations of frequentist methods, particularly regarding small sample sizes, heteroskedasticity, and endogeneity. The model assumes a Bernoulli distribution for the binary outcomes. Parameters were estimated using Markov Chain Monte Carlo (MCMC) simulations with a Gibbs Sampler algorithm (20,000 iterations, with a 2,500 iteration burn-in). The study reports posterior means, standard deviations, and posterior probabilities to determine the likelihood of effects.

Key Results
The Bayesian analysis yields the following probabilistic findings:

1. Positive Impact on Prosperity: Green accounting has a statistically significant positive effect on economic prosperity. The estimated coefficients range from 0.0042 to 0.0049, with posterior probabilities of occurrence between 99.92% and 99.99%. This suggests that transparency in environmental information and resource management supports economic growth.
2. Mitigation of Downturns: Green accounting exhibits a mitigating effect on economic downturns, with negative coefficients ranging from -0.0014 to -0.0017 and posterior probabilities from 97.78% to 99.99%. This indicates that GA helps stabilize the economy during negative cycles.
3. Asymmetric and Short-Term Effects:
   • While both the expansion and contraction of GA contribute to prosperity, these effects are noted as short-term.
   • Excessive Expansion: When GA is excessively expanded (beyond a reasonable threshold), its positive contribution to prosperity weakens (coefficient ~0.0002, probability ~67%). The authors suggest this may lead to information overload or reduced market confidence.
   • Severe Contraction: A severe lack of green accounting implementation (GAH2) significantly increases the risk of economic downturn (coefficient ~0.066, probability ~71.5%).
   • Excessive Expansion in Downturns: Paradoxically, excessive expansion of GA during downturns also worsens economic outcomes, suggesting that over-disclosure or poorly controlled implementation can create instability.

Key Contributions
The paper claims four primary contributions:

1. Theoretical: It establishes a theoretical foundation linking green accounting to both economic prosperity and downturns, moving beyond the view of GA solely as a sustainability tool.
2. Empirical: It provides evidence on the asymmetric effects of GA, specifically examining how excessive expansion and severe contraction differentially impact the economic cycle.
3. Methodological: By applying Bayesian Logistic Regression, the study addresses common econometric issues (autocorrelation, heteroskedasticity, endogeneity) and provides probabilistic predictions that are robust even with smaller samples or uncertain data, offering an alternative to traditional OLS or GMM approaches.
4. Policy-Oriented: It offers a composite measurement framework for GA, moving away from single-indicator metrics to a multidimensional index.

Significance and Claims
The authors assert that this study provides a nuanced perspective on the economic cycle, challenging the notion that "more" green accounting is always better. The significance lies in the finding that while GA is generally beneficial, its implementation must be balanced and well-targeted.

• For Policymakers: The results suggest that during economic expansions, governments should enhance transparency and integrate environmental costs into accounting systems to sustain growth. However, during downturns, while GA acts as a buffer, excessive implementation can be counterproductive.
• Strategic Shift: The paper advocates for a shift from merely expanding green accounting to refining it. It recommends integrating GA into national accounting systems to improve environmental management effectiveness while avoiding the "information noise" that can arise from over-disclosure.
• Limitations: The authors modestly acknowledge that the study does not separately analyze regional differences (e.g., Africa vs. Asia) or account for unexpected macroeconomic shocks. They suggest future research should narrow the scope by region and explore alternative measurement scales.

In conclusion, the paper argues that green accounting is a critical mechanism for reducing information asymmetry and corruption, thereby fostering sustainable growth, but its efficacy is contingent upon avoiding extreme levels of implementation that could destabilize the economy.