M E C H A N I S M · Energy Systems · Cross-Country Decarbonisation
1. Introduction
The global energy transition is often presented as a single curve bending towards low-carbon energy. At the country level, it looks very different. The diffusion of decarbonisation technologies is uneven, recent, and highly concentrated. A small group of countries is decarbonising rapidly, while many others are making only limited progress. In much of the world, renewable energy is expanding alongside fossil fuels rather than replacing them. Therefore, the global transition is not one transition but many, unfolding at different speeds and along different pathways.
This paper highlights that variation by establishing a set of stylised facts on the cross-country diffusion of decarbonisation technologies. Using public Energy Institute and World Bank data, I develop four reproducible measures that capture complementary dimensions of the transition: deployment level, decarbonisation velocity, carbon lock-in, and transition-path position.
2. What Are Measured?
I treat decarbonisation as a diffusion process and measure four complementary dimensions of that process. The first is deployment level, which captures how much of a country's energy system has already shifted towards low-carbon sources. The second is decarbonisation velocity, which measures how quickly that shift is taking place. Countries may reach similar deployment levels but arrive there at very different speeds. The third is carbon lock-in, which captures the extent to which existing fossil-fuel dependence resists change through established infrastructure, institutions, and patterns of energy use. The fourth is transition-path position, which identifies where a country sits along the broad sequence from coal- and oil-based systems towards gas and, ultimately, low-carbon energy.
3. Data and Methods
I use primary energy consumption by fuel from the Energy Institute Statistical Review of World Energy for the period 1965–2024. The dataset reports energy consumption in exajoules on a common energy basis. I combine oil, natural gas, coal, nuclear, hydroelectricity, solar, wind, and other modern renewables for each country and year. Country characteristics, including income classification, are obtained from the World Bank World Development Indicators. The final panel covers 79 countries reported individually in the Energy Institute dataset.
4. Stylised Facts
Fact 1. Diffusion is recent and concentrated
Modern renewable energy is a recent phenomenon for most countries. Across almost the entire sample period, its share of primary energy remained close to zero. Meaningful growth appears only after about 2000 and is concentrated in a relatively small group of countries. Even today's frontier economies show little evidence of sustained deployment before the turn of the century.
The diffusion curve suggested by global aggregates is, at the country level, a collection of late and uneven take-offs rather than a smooth and broadly shared transition. This distinction matters because countries are often compared as if they were moving along the same diffusion path. The evidence instead suggests multiple diffusion trajectories, with different starting points, speeds, and levels of technological adoption.
"The diffusion curve that the aggregate transition implies is, at the country level, a set of late and uneven take-offs rather than a smooth global shift."
Fact 2. Diffusion speed is stratified, and the frontier is pulling away
The pace of decarbonisation differs across countries. The distribution of decarbonisation velocity is highly concentrated near zero, with a long upper tail. The median country increases its modern renewable share by only about 0.19 percentage points per year, whereas a small group of frontier economies exceeds one percentage point annually. Most countries are therefore progressing slowly, while a limited number are advancing much more rapidly.
This unequal pace is reflected in the widening dispersion of low-carbon deployment. Between 1990 and 2023, the cross-country standard deviation of the low-carbon energy share increased from 4.0 to 9.7 percentage points, more than doubling over the period. Rather than converging towards a common transition pathway, countries are becoming increasingly differentiated in their levels of decarbonisation.
Fact 3. Carbon lock-in concentrates among stuck fossil economies
Carbon lock-in is not evenly distributed across countries. Instead, it is concentrated among a relatively small group of economies that combine persistently high fossil-fuel dependence with little evidence of structural change. Trinidad and Tobago, Turkmenistan, Kuwait, Iraq, and Algeria represent the strongest examples. Their energy systems remain overwhelmingly dependent on fossil fuels, with little sustained decline over the observation period.
At the opposite end of the distribution are countries that have already diversified their energy systems through hydroelectricity, nuclear power, or rapid renewable deployment. The evidence suggests that carbon lock-in is not a universal characteristic of the global energy transition but a feature of a distinct group of 'stuck' economies.
Fact 4. Addition, not substitution
In many countries, renewable energy has expanded without reducing fossil-fuel consumption. Between 2010 and 2023, 46% of the countries in the sample (36 of 79) increased their modern renewable share while simultaneously increasing the absolute volume of fossil-fuel consumption. China provides the clearest example. Its modern renewable share increased by more than 4.4 percentage points, yet fossil-fuel consumption rose by 42 exajoules over the same period. India, Indonesia, and Vietnam display a similar pattern.
These countries are undergoing energy addition rather than energy substitution. Renewable energy is meeting part of rapidly growing energy demand, but it is not yet replacing the existing fossil-fuel base.
5. Discussion
Taken together, the four stylised facts describe a global energy transition that is late, uneven, and increasingly stratified. Modern renewable deployment is concentrated in a relatively small group of countries. Decarbonisation is progressing rapidly at the frontier but slowly across most of the world. Carbon lock-in remains concentrated among a distinct group of fossil-dependent economies, while renewable expansion frequently occurs alongside continued growth in fossil-fuel consumption. Thus, the global transition appears less as a single accelerating process than as multiple national transitions following different trajectories.
6. Conclusion
The global energy transition is not a single curve but a collection of national transitions unfolding at very different speeds. Here, I measured the diffusion of decarbonisation technologies across countries along four complementary dimensions. I showed that it is recent, stratified, diverging, and, in much of the world, additive rather than substitutive. Together, these measures provide a consistent basis for comparing national decarbonisation trajectories and identifying the empirical regularities that future research should explain.
References
[1] Grubler, A. (2012). Energy transitions research: Insights and cautionary tales. Energy Policy, 50, 8–16.
[2] Sovacool, B.K. (2016) How Long Will It Take? Energy Research & Social Science, 13, 202–215.
[3] Rogers, E. M. (2003). Diffusion of Innovations (5th ed.). Free Press.
[4] Islam, N. (2003). Journal of Economic Surveys, 17: 309–362.
[5] Barro, R., & Sala-i-Martin, X. (1992). Journal of Political Economy, 100, 223–251.
[6] Unruh, G.C. (2000). Understanding carbon lock-in. Energy Policy, 28(12), 817–830.