📈 economics

Whose energy transition? Global renewable-capacity resilience masks country-level divergence after shocks

While global renewable energy capacity has demonstrated resilience against major economic and geopolitical shocks, this aggregate growth masks significant country-level divergence, as the typical nation's trajectory often deviates from the global trend and is heavily influenced by a few large economies.

Original authors: Saumitra Saxena, Bassam Dally

Published 2026-07-01

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Saumitra Saxena, Bassam Dally

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: The "Class Average" vs. The "Typical Student"

Imagine a school where the total number of books in the library has grown steadily every year, even during bad times like a pandemic or an economic crash. If you just look at the total number of books, you might think, "Wow, everyone is reading more!"

But this paper asks a different question: "Is the typical student actually reading more, or is just one or two super-wealthy students buying thousands of books and skewing the total?"

The authors, Saumitra Saxena and Bassam Dally, looked at the world's renewable energy (solar and wind power) like this school library. They wanted to see if the global growth story matches the reality for individual countries, especially when the world faces big shocks like financial crises, oil price crashes, or wars.

The Four "Storms" They Studied

The researchers looked at four specific "storms" that hit the global economy:

2008: The Global Financial Crisis.
2014: The Oil Price Crash (oil got very cheap).
2020: The COVID-19 Pandemic.
2022: The Russia-Ukraine War.

They compared what actually happened to what should have happened if countries had just kept growing at their usual speed.

The Main Findings

1. The Global Total is a "Trick of the Light"

When you add up every country's solar and wind power, the total keeps going up. It looks like the world is resilient. However, the paper shows that this "global average" is mostly driven by a few giant economies (like China, the US, and Germany).

The Analogy: Imagine a classroom where the average height is 6 feet. That's because two basketball players are 7 feet tall. If you remove them, the "typical" student is actually much shorter. Similarly, the world's renewable energy growth is being propped up by a few massive countries.

2. The 2014 Oil Crash Was the Real "Brake"

Among the four storms, only one actually caused a widespread slowdown for the "typical" country: The 2014 Oil Price Crash.

What happened: When oil became very cheap, many countries that were already building solar and wind farms slowed down their construction.
The Evidence: The authors used a "placebo test" (like a control group in a science experiment). They checked if countries slowed down during normal years. They didn't. But in 2014, the typical country's growth slowed down significantly compared to the norm.
The Catch: This slowdown was mostly seen in countries that were already building a lot of renewable energy. Countries that hadn't started yet didn't show this effect.

3. The Other Storms Were Different

2008 & 2020: The global total looked fine, but the "typical" country actually slowed down a bit. However, the big economies (like China and the US) kept building so fast that they covered up the slowdown for everyone else. It was like a few rich students buying enough books to hide the fact that the rest of the class stopped reading.
2022 (War): The global total dropped a bit. But when the researchers removed the biggest countries from the calculation, the drop looked much worse. This means the biggest economies actually helped cushion the blow for the global total, while smaller countries with big plans suffered more.

4. Why Did This Happen? (The "Why" Question)

The authors tried to find out why some countries slowed down and others didn't. They looked at income levels and whether a country sells oil.

The Result: They found some hints (like oil-exporting countries might slow down more when oil is cheap), but the evidence wasn't strong enough to be certain.
The Mechanical Factor: The biggest predictor of a slowdown was simply how fast a country was growing before the shock. If you were growing very fast, you were more likely to hit a "speed bump" when the economy got shaky.

The Takeaway

The paper concludes that the world's renewable energy transition is real, but uneven.

The Headline: "Global renewable capacity is growing!"
The Reality: "The global total is growing, but for many individual countries, the growth slowed down or stopped during major crises. A few giant countries are doing the heavy lifting, masking the struggles of the rest."

The Final Metaphor:
Think of the global renewable energy transition as a marathon. The finish line (the total capacity) is moving forward. But if you look at the runners, a few elite athletes are sprinting so fast they are pulling the average time down. Meanwhile, the "typical" runner is stumbling or slowing down when the weather gets bad (economic shocks). The paper warns us not to just look at the finish line; we need to look at the runners to see who is actually struggling.

Technical Summary: "Whose energy transition? Global renewable-capacity resilience masks country-level divergence after shocks"

Problem Statement
While global renewable energy capacity (solar plus wind) has demonstrated aggregate resilience, continuing to grow through major economic and geopolitical shocks (2008 GFC, 2014 oil-price crash, 2020 COVID-19, 2022 Russia–Ukraine war), the global total may not represent the typical country. The aggregate trajectory is heavily dominated by a small set of national programs, particularly China, which accounted for nearly half of global capacity by end-2024. The paper addresses the divergence between aggregate trends and country-level outcomes, questioning whether global resilience masks significant contractions or deviations in specific nations. It seeks to quantify these deviations for four major shocks and assess the robustness of aggregate findings against sample selection, trend extrapolation, and the influence of large economies.

Methodology
The study utilizes a panel of 222 countries with installed solar and wind capacity data from 2000 to 2024. The core analysis focuses on four shock years: $T \in \{2008, 2014, 2020, 2022\}$ .

Sample Selection: The primary sample includes countries with at least 50 MW of solar plus wind capacity in the year prior to the shock ( $T-1$ ), focusing on "established adopters." A secondary specification uses an inverse-hyperbolic-sine (IHS) transformation to include smaller adopters (at least two non-zero pre-shock observations).
Gap Estimation: For each country, a log-linear trend is fitted to non-zero observations in the four-year pre-shock window ( $T-4$ to $T-1$ ), requiring at least three data points. This trend is projected to $t+0, t+1,$ and $t+2$ . The "gap" is defined as the percentage deviation of observed cumulative capacity from the projected capacity. A negative gap indicates slower growth than projected, not necessarily a decline in absolute capacity.
Distributional Measures: Three metrics are calculated for each shock:
1. Aggregate Gap ( $G$ ): The capacity-weighted mean of country gaps.
2. Unweighted Mean ( $\mu$ ): The simple average of country gaps.
3. Country Median ( $m$ ): The median of country gaps, reported with 95% percentile-bootstrap confidence intervals.
Placebo Benchmark: To distinguish shock effects from noise in short-window extrapolations, the study compares real-shock medians against empirical placebo bands derived from non-shock years (2004, 2005, 2011, 2017). Two specifications are used: a strict two-year buffer and a "window-clean" specification.
Robustness Checks:
- Leave-out Tests: The aggregate gap is recalculated excluding China and the top 3 or 5 economies (ranked by pre-shock capacity) to test if large economies drive the aggregate.
- Exploratory Regression: OLS regressions (with shock and horizon fixed effects) test correlates of country-level gaps, including pre-shock trend slope, log capacity, oil-exporter status, income groups, and oil rents.
- SVAR: A supplementary sign-restricted Structural Vector Autoregression (SVAR) is performed to check for structural responses in global generation.

Key Results

Divergence in the 2014 Oil-Price Crash: Among established adopters, the 2014 oil-price crash is the only shock where the country median gap falls outside the placebo bands in the contraction direction across all post-shock horizons ( $t+0, t+1, t+2$ ). For example, at $t+0$ , the median gap was $-11.2\%$ , significantly below the strict placebo band of $[-9.1\%, -1.8\%]$ . This result is sample-sensitive; it disappears when the IHS specification admits smaller adopters, suggesting the deviation is concentrated among countries with material project pipelines.
Aggregate vs. Country-Level Divergence:
- 2008 and 2020: The aggregate gap ( $G$ ) was near zero, while the unweighted mean ( $\mu$ ) was significantly negative (e.g., $G = -0.3\%$ vs. $\mu = -7.9\%$ in 2008). This indicates that countries with larger projected capacities experienced less negative gaps, buffering the aggregate.
- 2022: The aggregate gap was negative ( $-8.0\%$ ), but the median was only slightly negative ( $-1.5\%$ ). Leave-out tests reveal that excluding the top five economies shifts the aggregate gap from $-8.0\%$ to $-18.8\%$ , indicating that the largest economies (including Germany) attenuated the aggregate deterioration, while other substantial economies faced deeper contractions.
- 2014: Both the aggregate and median were negative, showing a broadly distributed deviation without the concentration seen in other years.
Correlates of Deviation: Exploratory regressions find that the fitted pre-shock trend slope is the dominant correlate of the gap (partly mechanical, as it defines the projection). Evidence for income levels and oil-exporter status is suggestive but not robust across specifications or significance thresholds (e.g., income $p=0.218$ at $t+0$ ).
SVAR Findings: The supplementary SVAR exercise finds no 95%-distinguishable renewable-generation response to the identified macroeconomic or geopolitical shocks.

Significance and Claims
The paper claims that global renewable-capacity resilience is real but unevenly distributed. The aggregate trajectory often masks significant country-level divergence, particularly for established adopters facing specific shocks like the 2014 oil-price crash.

Methodological Contribution: The study demonstrates how relationships between shocks and renewable deployment appear across the country distribution rather than as a single panel coefficient. It highlights the limitations of aggregate targets in capturing the heterogeneity of the transition.
Policy Implication: Because global ambitions (e.g., tripling renewable capacity) are framed as aggregate targets, progress reports must also show deployment distribution. Aggregate expansion can coexist with uneven participation or contraction in specific subsets of countries.
Risk Assessment: The findings suggest that prolonged low-oil-price conditions may pose a greater risk to renewable deployment in countries with mature project pipelines (where the 2014 effect was observed) compared to countries that have not yet begun material deployment.
Caveats: The authors explicitly state that the design is descriptive and does not identify causal treatment effects. The 2014 result is sample-sensitive and should not be generalized to countries without pre-shock build-outs. The 2026 Iran-war episode is discussed only as context, with the historical results serving as risk markers for future testing rather than forecasts. The study identifies candidate correlates for future work rather than definitive policy levers.