Countries Reducing Energy Intensity Fastest, 2015–2022
Which economies reduced energy use per unit of GDP fastest?
Liberia recorded the fastest decline in energy intensity between 2015 and 2022, followed by the Republic of the Congo and Seychelles. Their compound annual reductions were 8.55%, 7.33% and 6.65%, respectively.
Energy intensity is the amount of primary energy used to generate one international dollar of inflation-adjusted, purchasing-power-adjusted GDP. A falling value means that an economy required less energy for each unit of measured output at the end of the period.
Dataset version: Our World in Data energy dataset dated April 27, 2026. Calculation checked on July 16, 2026. Full commit SHA: 7e387a16f70a510e433f8aac7efeac6faa1e5059.
Direct answer: Liberia ranked first with an 8.55% yearly compound decline. The Republic of the Congo ranked second at 7.33%, while Seychelles ranked third at 6.65%. The ranking measures historical change from 2015 to 2022; it is not a forecast for 2025 or 2026.
Liberia ranked first using values before display rounding.
Liberia’s 2022 value was 46.51% below its 2015 level.
Rows with an ISO code and values in both 2015 and 2022.
Kilowatt-hours of primary energy per international dollar of GDP.
What the metric measures
The source field, energy_per_gdp, measures primary energy consumption in kilowatt-hours per international dollar of GDP. The GDP denominator is adjusted for inflation and differences in living costs between economies.
The table converts the change between 2015 and 2022 into a compound yearly rate. A result of 5% does not mean that the value fell by exactly 5% during every calendar year. It means that a constant decline of 5% per year would produce the same difference between the two selected years.
Higher positive rates rank higher. Economies where energy intensity increased remain outside the displayed Top 20 because the table is ordered by the largest declines.
How to read the result: a larger positive percentage indicates a faster decline in energy used per unit of GDP.
What it does not prove: the metric is not a direct measure of factory efficiency, climate-policy quality or emissions performance.
Why results differ: industrial structure, services, climate, energy supply and changes in measured GDP can all influence the ratio.
Top 20 countries and economies, 2015–2022
The table ranks the leading countries and economies by the compound yearly decline in energy intensity. Search by name, filter by region or change the display order without altering each economy’s original rank.
Compound yearly decline in energy intensity from 2015 to 2022
| Rank | Economy | Yearly decline | 2015 → 2022 |
|---|---|---|---|
| 1 | Liberia LBR | 8.55% | 1.260 → 0.674 kWh/int-$ Total decline: 46.51% |
| 2 | Republic of the Congo COG | 7.33% | 0.939 → 0.551 kWh/int-$ Total decline: 41.32% |
| 3 | Seychelles SYC | 6.65% | 1.912 → 1.181 kWh/int-$ Total decline: 38.23% |
| 4 | Guinea-Bissau GNB | 6.11% | 0.689 → 0.443 kWh/int-$ Total decline: 35.70% |
| 5 | Hong Kong SAR, China HKG | 6.02% | 0.942 → 0.610 kWh/int-$ Total decline: 35.24% |
| 6 | Djibouti DJI | 5.88% | 1.498 → 0.980 kWh/int-$ Total decline: 34.58% |
| 7 | Cabo Verde CPV | 5.88% | 1.073 → 0.702 kWh/int-$ Total decline: 34.58% |
| 8 | Gambia GMB | 5.25% | 0.658 → 0.451 kWh/int-$ Total decline: 31.46% |
| 9 | Cuba CUB | 4.83% | 1.442 → 1.020 kWh/int-$ Total decline: 29.26% |
| 10 | Romania ROU | 4.59% | 0.992 → 0.714 kWh/int-$ Total decline: 28.02% |
| 11 | New Zealand NZL | 4.22% | 1.569 → 1.160 kWh/int-$ Total decline: 26.07% |
| 12 | Estonia EST | 4.20% | 2.137 → 1.583 kWh/int-$ Total decline: 25.92% |
| 13 | Slovenia SVN | 4.18% | 1.373 → 1.018 kWh/int-$ Total decline: 25.86% |
| 14 | Benin BEN | 4.13% | 1.022 → 0.761 kWh/int-$ Total decline: 25.54% |
| 15 | Nicaragua NIC | 4.07% | 0.876 → 0.655 kWh/int-$ Total decline: 25.23% |
| 16 | France FRA | 3.96% | 1.145 → 0.863 kWh/int-$ Total decline: 24.63% |
| 17 | Gabon GAB | 3.85% | 0.578 → 0.439 kWh/int-$ Total decline: 24.05% |
| 18 | Uzbekistan UZB | 3.78% | 1.971 → 1.505 kWh/int-$ Total decline: 23.64% |
| 19 | São Tomé and Príncipe STP | 3.64% | 0.914 → 0.705 kWh/int-$ Total decline: 22.87% |
| 20 | Luxembourg LUX | 3.59% | 1.389 → 1.075 kWh/int-$ Total decline: 22.61% |
Ranks are calculated from values before rounding. Changing the display order does not assign new ranks.
How the leading rates compare
Liberia’s 8.55% result is used as the full bar width. The remaining bars show each economy’s rate relative to the leader.
How the results were calculated
Period
2015 is the common starting year selected for this analysis. The latest common year available for the source field is 2022.
Included rows
Countries and economies with an ISO alpha-3 code and a value in both selected years.
Excluded rows
Regional totals, income groups, continents and rows missing either the 2015 or 2022 value.
Order
The largest positive compound decline ranks first. Calculations use the source values before display rounding.
Seven is used because the interval from 2015 to 2022 contains seven year-to-year periods. The selected starting year creates a consistent interval for every included economy; it is not presented as an official starting year required by the source.
The source values come from energy_per_gdp in the Our World in Data energy dataset. The field combines published energy and GDP inputs to express primary energy consumption in kilowatt-hours per international dollar.
Coverage check: the source file was filtered to rows with an ISO alpha-3 code, then restricted to economies with non-missing values in both 2015 and 2022. Aggregates and incomplete rows were excluded. The resulting set contains 165 countries and economies.
No missing values were estimated, and no figures for 2023–2026 were projected. The method compares the first and last year of the period; it does not show whether progress was steady, concentrated in one year or interrupted by temporary reversals.
Patterns visible in the Top 20
Clear leader
Liberia’s 8.55% rate is 16.6% higher than the Republic of the Congo’s 7.33% result when the difference is measured relative to the second-ranked rate.
Small leading group
Only three economies exceed 6.5%. The rate falls below 6% after Hong Kong SAR, China.
Regional distribution
Sub-Saharan Africa contributes eight entries and Europe and Central Asia contributes six. Together they represent 70% of the Top 20.
Close results
Djibouti and Cabo Verde both display 5.88%, but Djibouti ranks higher using the values before rounding. Estonia and Slovenia are separated by only 0.013 percentage points.
The table alone does not establish why a particular country improved. Changes in industry, services, energy supply, tourism, weather or measured GDP are possible influences on the ratio, but country-specific causes require national and sector-level evidence.
Results can also depend on the selected years. A disruption affecting either 2015 or 2022 may change the compound rate, so the full annual series should be reviewed before drawing policy conclusions.
How to interpret differences between countries
This comparison identifies where primary energy use per unit of purchasing-power-adjusted GDP declined fastest over the selected period. It can help identify countries that merit closer analysis of industrial change, energy technology, buildings, transport and economic structure.
A high position does not automatically mean that total energy consumption fell. Energy intensity can decline while total demand rises when economic output grows faster than energy use.
The metric also does not provide a direct comparison of individual factories, vehicles or buildings. Economies dominated by mining, refining or heavy manufacturing generally use more energy per unit of GDP than economies with a larger share of finance, software and other services.
Frequently asked questions
Are these 2025 or 2026 values?
No. The measured period is 2015–2022. The dataset version is dated April 27, 2026, and the calculations were checked on July 16, 2026.
Why are some countries missing?
A country or economy is excluded when the source does not provide an ISO-coded row with values in both 2015 and 2022. Regional totals and other aggregates are also excluded.
Why is Hong Kong included?
The scope covers countries and economies rather than sovereign states only. Hong Kong SAR, China has its own ISO alpha-3 code and complete values for both selected years.
Does a 5% result mean a 5% fall every year?
No. It is a compound rate calculated from the first and last year. Actual yearly changes may have been uneven.
Is this the official SDG 7.3.1 ranking?
No. This page applies a stated formula to a documented OWID field. Official SDG reporting should be read from the responsible international agencies.
Can total energy consumption rise while energy intensity falls?
Yes. This happens when GDP grows faster than primary energy consumption.
Does a high rank prove successful energy policy?
No. Policy may contribute, but economic restructuring, energy supply, weather, disruptions and statistical revisions can also affect the result.
Sources and data documentation
Our World in Data energy CSV
Exact file used for the 2015 and 2022 values in the table.
https://raw.githubusercontent.com/owid/energy-data/7e387a16f70a510e433f8aac7efeac6faa1e5059/owid-energy-data.csvOur World in Data energy codebook
Definitions, units and source information for energy_per_gdp.
Our World in Data commit record
Repository record for the dataset version dated April 27, 2026.
https://github.com/owid/energy-data/commit/7e387a16f70a510e433f8aac7efeac6faa1e5059United Nations SDG 7.3.1 metadata
International methodology and limitations for interpreting energy intensity.
https://unstats.un.org/sdgs/metadata/files/Metadata-07-03-01.pdfWorld Bank energy intensity indicator
Reference page for the international primary-energy-intensity indicator relative to PPP GDP.
https://data.worldbank.org/indicator/EG.EGY.PRIM.PP.KD
