Discussant Comments�Session VI. Climate Finance and Fiscal Policy �Paper: Climate Finance, Renewable Energy, and Fiscal Sustainability in Pacific Island Developing States (PIDs)�Author: Homer Pagkalinawan

�Discussant: Dr. Sabah Abdulla, Senior Economist (Climate Change), ADB

�Climate Change Fiscal Policy Conference 2, Tokyo, Japan

10 July 2025

INTERNAL. This information is accessible to ADB Management and Staff. It may be shared outside ADB with appropriate permission.

Recap on the Paper

• An important contribution to the Small Island Developing States (SIDS) literature, particularly in the Pacific Island Developing States (PIDS) context, by exploring the role of renewable energy in reducing fossil fuel dependency.
• The paper adopts a mixed-methods approach, combining qualitative analysis of Nationally Determined Contributions (NDCs) with a fixed effects panel regression covering the period 2000–2022 for 170 countries. The econometric model controls for a range of factors, including governance effectiveness, renewable energy share, energy intensity, and oil prices.
• The results show that governance effectiveness positively mediates the relationship between renewable energy investment and reduced fossil fuel dependency.
• The comparative NDC mapping analysis is particularly valuable and region-specific—similar in spirit to the work we’ve done in climate adaptation financing through National Adaptation Plans (NAPs) see the APCR 2024, chapter 3 on Adaptation.

​

INTERNAL. This information is accessible to ADB Management and Staff. It may be shared outside ADB with appropriate permission.

Key Questions & Technical Considerations

1. Theoretical Underpinning: What is the central causal question? Is the paper identifying the drivers of fossil fuel dependency, or the barriers to renewable energy adoption?

• A clearer theoretical model is needed to frame the analysis appropriately.

​

2. Model Selection and Justification: The paper uses fixed effects (FE) but does not justify this choice beyond mentioning “R-squared” and “theoretical consistency.”

• Include a Hausman test or provide a stronger theoretical rationale for preferring FE over pooled OLS or random effects (e.g., country-specific unobserved heterogeneity).

​

3. Endogeneity Concerns: Potential for reverse causality (e.g., high fossil fuel dependency → more climate finance) and omitted variable bias is not addressed. Claim in the paper “Climate finance and governance reduce fossil fuel dependency.” BUT Potential Reverse Causality Concern: Does fossil fuel dependency cause more climate finance to be allocated?

• Acknowledge limitations and consider using IV approaches or lagged variables to mitigate endogeneity. Also, it’s important to clarify that the paper uses renewable energy investment as a proxy for climate finance, rather than measuring climate finance flows directly

​

4. Non-Linearity: Inverted-U Specification: Paper proposes an inverted-U relationship between renewable energy share and fossil fuel imports. However, no theoretical justification or prior empirical reference is provided.

• Link this to energy transition, adoption curves, or cite similar findings (e.g., Environmental Kuznets Curve-type dynamics).

​

5. Mis-specification of the Governance Effectiveness Variable: The interaction term “RE × Governance” is included in the regression, but the main effect of Governance Effectiveness is omitted in some models. Omitting the main effect can lead to incorrect interpretation of the interaction term and overall model mis-specification.

• Ensure all interaction terms are accompanied by their respective main effects in all specifications.

.

INTERNAL. This information is accessible to ADB Management and Staff. It may be shared outside ADB with appropriate permission.

Final Comments, Suggestions & Reflections

1. Logical Framing and Structure: Two framing options need clearer articulation:

1. Idea 1 – Fossil Fuel Dependency + Transition Risks (NDC focus): While the paper usefully links fossil fuel dependency and renewable energy targets found in PIDS’ NDCs, it misses a discussion of transition risks—especially those related to demand shifts, technology adoption, and policy uncertainty.
1. Consider: Integrating broader theoretical underpinnings from the literature on climate finance and transition risks, focusing on drivers of RE investments, rather than just system reliability (SAIDI). OR

b. Idea 2 – Climate Finance as a Driver of Renewable Energy Investment:: An alternative framing could emphasize the pathway:�Climate finance → RE investments → reduced fossil fuel imports → fiscal sustainability.�If climate finance is the focus, the paper should draw on actual climate finance flows (e.g., concessional, non-concessional, or grant-based ODA from datasets like OECD-CRS or CPI), rather than treating RE investment as a proxy.

• Consider: Reframing the dependent variable to focus on renewable energy investments or RE generation share, with fossil fuel import dependency as a consequence or control variable, not the main outcome.

2. Econometric Strategy and Endogeneity

a. Model Enhancement Needed: The paper uses fixed effects models but does not adequately address endogeneity, which may arise due to joint determination of RE investments and fossil fuel dependency by omitted factors (e.g., climate finance inflows, governance, aid absorption capacity).

• Consider: Applying 2SLS with lagged instruments, or use exogenous instruments such as: Historical aid flows (e.g., GCF, ODA), Commodity price shocks (e.g., oil price volatility). Reference: See the Recent work by Salah’Uddin et al. (2025) provides a more rigorous framework using PCSE, 2SLS, and quantile regression to capture distributional heterogeneity and reduce simultaneity bias. The paper could benefit from adopting similar techniques, particularly: Lagging RE variables, Testing for cross-sectional dependence, and Using panel quantile regression to distinguish low- and high-investment country trajectories.

b. Include Summary Statistics and Theoretical Mapping: Depending on which framing is adopted (Idea 1 or 2), the paper would benefit from a clear summary statistics table and stronger linkage to theoretical constructs like climate-related risk types or vulnerability

3. NDC Mapping Strategy: The fiscal dimensions analyzed in the PIDS' NDCs—such as fiscal mechanisms, quantified costing, private sector mobilization—are important, but the analysis may suffer from selection bias if limited to only 14 countries.

• Consider: Using a comparative framework (as in the regression model) to map these fiscal dimensions across a larger global NDC dataset (e.g., 100+ countries). This would improve external validity and allow cross-country comparisons of fiscal innovation and climate ambition.

4. Final Reflections: ADB’s operations relevance- This paper is a valuable contribution to ADB’s evolving work on fiscal-climate integration and energy transition in small island contexts.

• Moving forward: Addressing econometric limitations (via 2SLS or dynamic panel models) and strengthening the causal logic around climate finance → RE investment would significantly enhance both the empirical robustness and theoretical credibility of the study.

INTERNAL. This information is accessible to ADB Management and Staff. It may be shared outside ADB with appropriate permission.