Work In Progress
Resource Dependence and the Allocation of International Clean Energy Finance: Evidence from Developing Countries (Work in Progress)
Abstract: International clean energy finance does not reach developing countries evenly. A small group of recipients captures most of the flows. This paper asks whether natural resource dependence shapes that pattern and whether donors read different resource endowments as different signals. The analysis uses a balanced panel of 57 developing countries from 2000 to 2022. The sample covers about 72% of total committed clean energy finance. The empirical strategy is IV-GMM with Bartik shift-share instruments. These are built from country-specific resource exposure and global commodity prices. What the results show is that the effect of resource dependence depends on the type of resource. Fossil fuel rents link to lower inflows, which fits the carbon lock-in story. Mineral rents link to higher inflows, but only in countries producing minerals that matter for the energy transition. This points to donor interest in securing supply chains. For conventional mineral producers, the link is negative and mirrors the fossil fuel pattern. The same patterns show up in grant-based flows. So donor decisions drive the distribution, not commercial risk alone. Governance attracts finance but does not offset the penalty tied to fossil fuel dependence. Countries that rely most on carbon-intensive development need transition support the most and find it the hardest to get.
Labor Participation and Renewable Adoption Across Developing Countries: Identification and Mechanisms (Under Review at Review of Development Economics )
Co-authored with Ankit Singh Kharwar
Abstract: This paper examines how labor force participation influences the energy transition in developing economies, focusing on the shift from fossil fuels to renewable energy. Using an unbalanced panel of 70 countries from 2002–2021, we estimate fixed-effects and two-step GMM models with Driscoll–Kraay standard error to address heterogeneity, cross-sectional dependence, and endogeneity. The results show that higher labor force participation significantly reduces fossil fuel consumption and increases renewable energy uptake. Mediation analysis identifies regulatory quality as a channel through which labor participation promotes renewables, while moderation tests show that government effectiveness and rule of law amplify this effect. Robustness checks confirm the causal interpretation. These findings highlight labor force dynamics as an underexplored determinant of national energy portfolios. Policy implications point to labor market reforms, skills upgrading, and governance improvements as levers to accelerate renewable adoption in developing countries.
Does Fiscal Space Shape Access to International Clean Energy Finance? Evidence from 32 Developing Countries (Under review at International Review of Economics & Finance)
Co-authored with Deepak Deepak
Abstract: The financing gap for renewable energy in developing countries is widely acknowledged, yet the role of fiscal space in shaping access to international clean energy finance remains underexplored. Existing studies have examined financial development, green finance tools, and climate-related fiscal mechanisms, but none have empirically tested how a country's fiscal position influences its ability to attract international public finance for clean energy. This study addresses this gap using panel data from 32 developing countries over 2000–2021. The analysis employs fixed-effects estimation with Driscoll-Kraay standard errors and IV-GMM to address endogeneity and cross-sectional dependence. The results reveal a positive and robust association between fiscal space, proxied by net lending/borrowing, and international clean energy finance. Countries with lower GDP per capita receive more support, while renewable energy consumption does not show a significant effect. These findings suggest that stronger fiscal positions may improve donor confidence and signal investment readiness. From an environmental management perspective, fiscal capacity emerges as a foundational condition for scaling low-carbon technologies and advancing progress toward SDG~7 (affordable and clean energy) and SDG~13 (climate action).
Policy Instrument Design and the Durability of Transport Climate Policy (Work in Progress)
Abstract: Transport decarbonization depends not only on effective policies but also on whether they remain in force long enough to take effect. This study examines the durability of transport climate policies, a dimension that has received limited attention. Using the Climate Policy Database and measures of institutions, governance, and development, the study applies event-history analysis to 1,304 national transport climate policies across 91 countries between 1990 and 2025. Policies are classified by principal instrument as binding, fiscal, or soft. Kaplan-Meier and Cox proportional-hazards models show a durability gradient across instrument types. Fiscal policies face approximately three times the discontinuation hazard of binding policies, robust across specifications and termination definitions. Competing-risks analysis further shows that binding policies are more often replaced or upgraded, whereas fiscal policies are more often abandoned. These findings indicate that instrument choice is partly a choice about policy longevity and that policy inventories may overstate durable climate commitment.
From policy to delivery: Verifying heavy-duty electrification policy from India's public vehicle register (Work in Progress)
Abstract: Governments increasingly commit to electrifying heavy-duty road transport, and expect those commitments to be checked against the public registers that record vehicle registration. This paper asks what such a register can establish about the delivery of sub-national electric-vehicle policy, using India, where states issued first policies on staggered dates between 2018 and 2023. Reconstructing the complete national record, 36 jurisdictions and 96 months holding 10,597 electric and 2,316,764 all-fuel heavy-duty registrations with no imputation, neither a Poisson nor a Callaway–Sant’Anna estimator detects an effect of policy activation, and the design could only have detected a 5.6-fold change. The record cannot separate a policy of that size or smaller from one with no effect. Retrieved separately, the vehicle categories show the outcome is 94.5 per cent buses, so aggregate estimates describe bus procurement rather than freight. One reporting convention compounds the difficulty and travels beyond India: the register suppresses zero rows, which makes a jurisdiction’s visibility a function of its own outcome. A screening rule requiring returned rows on both sides of a policy date then removes jurisdictions chiefly on the entry margin. Seven of the thirteen it discards had not yet entered the market, among them Delhi, which holds 24.4 per cent of national electric heavy-duty registrations. The twelve-month comparison gives 0.54 across the four retained and 1.41 across the thirteen removed, the retained figure resting on one jurisdiction reverting from an unusually active pre-window. The paper specifies five dated public records that would make heavy-duty policy delivery verifiable.
Crop-output safeguards and agrivoltaic investment in ten countries across Africa and South Asia
(Work in Progress)
Abstract: Agrivoltaic land-use policy must consider whether profitable solar investment also retains agricultural output. This study examines how crop-output requirements alter electricity production, investment returns and indicative public support in ten countries across Africa and South Asia. A common cash-flow model uses 10,000 parameter draws and paired comparisons on the same original hectare. Baseline median electricity costs net of crop-revenue changes range from 0.084 to 0.125 USD/kWh. Crop-revenue changes affect these costs by at most 2.4% at the median, yet maize output falls to approximately half its open-field level. Paired scenarios impose 80% and 90% whole-hectare output floors, each satisfied in at least 90% of the specified model draws. With land loss held fixed, the maize coverage caps retain about 24% and 2% of baseline electricity generation. Allowing land loss to decline with coverage raises these shares to about 32% and 16%. Fixed costs per hectare can raise costs and support needs per unit of installed capacity as coverage falls. Financing and electricity value also shape viability. The analysis links agricultural-output requirements to their investment consequences, extending existing work on agrivoltaic profitability and farmland preservation. It supports joint appraisal of crop retention and energy finance before designing public support. Transferred crop responses, price proxies and assumed redesign costs limit interpretation; the scenarios do not establish locally feasible designs, crop guarantees or optimal safeguards.