International efforts to map trajectories towards the achievement of sustainable development goals generally acknowledge the complementarity of renewable energy deployment and energy efficiency measures.1 For example, in 2011 the United Nations’ Sustainable Energy for All (SEforALL) initiative, recognising the joint role of renewables and efficiency in securing universal access to sustainable energy, set a target to double both the share of renewable energy in global final energy consumption (to 36% by 2030) and the rate of improvement in energy efficiency.2

Subsequently, in the second year of the United Nations Decade of Sustainable Energy for All (2014-2024), the UN General Assembly adopted the 2030 Agenda for Sustainable Development and the Agenda’s 17 Sustainable Development Goals (SDGs).3 SDG 7, on ensuring sustainable energy access, reaffirmed the earlier SEforALL objectives but replaced the 36% renewables target with a more general goal to “increase substantially the share of renewable energy in the global energy mix” by 2030.4

In 2018, the Intergovernmental Panel on Climate Change presented several pathways for mitigating climate change that are consistent with a relatively high probability of limiting the long-term increase in global average temperature to 1.5 degrees Celsius above pre-industrial levels.5 Each of the outlined pathways is characterised, in part, by relative reductions in global energy demand.6 Reducing energy demand requires advances in both energy efficiency (technology-specific) and energy conservation (behaviour-specific).7

Efforts have been made to disaggregate the effects of the three main determinants of total final energy demand: structural changes within economies, changes in the level of activity in each economic sector and changes in the efficiency of energy use in each sector.8 Such analysis has indicated that, without improvements in energy efficiency, global final energy demand in 2017 would have been 12% higher than 2000 levels.9 This translates to an average annual displacement of energy demand of below 0.7% during this period. Meanwhile, between 2005 and 2017, the share of total final energy consumption (TFEC) met by modern renewables grew at an average annual rate of 2.9%.10

In 2017, however, the estimated total renewable share in TFEC (modern renewables and traditional biomass combined) increased by merely 0.8%.11 ( See Global Overview chapter.) That same year, global energy demand rose an estimated 1.9% as the economy grew 3.7%, reflecting a decline in overall energy intensityi of 1.7% – the slowest rate of improvement since 2010.12 The linkages between improvements in energy efficiency and growth in renewable energy are evident.13 For example, the growth in non-thermal renewable energy improves primary energy intensity, and the improvement in end-use energy efficiency may stimulate further economic activity and, in turn, additional deployment of renewables.14

Government policies are instrumental in improving energy efficiency in the end-use sectors of buildings, industry and transport. For example, policies supporting energy efficiency in the European Union (EU) have been credited with advancing the share of renewable heat in buildings to 22% in 2017, as the demand for heat in the region stabilised and dipped slightly between 2012 and 2017 (-0.3%), making the EU the only region in the world where heat demand is declining.15

Worldwide, policy support for energy efficiency increased substantially between 2010 and 2017.16 However, the greatest advance was in the proliferation of national energy efficiency action plans and targets, whereas the number of specific national mandates grew more slowly.17 City governments adopted building energy codes, minimum energy efficiency performance standards and other firm commitments, and cities continued to play a prominent role in designing and implementing policies for energy efficiency. ( See Feature chapter.)

Although both renewables and efficiency are critical elements of more sustainable energy systems, policy makers may struggle with where to allocate resources most effectively: on the supply side (renewable energy) or on the demand side (energy efficiency). In the United States, the proliferation of renewable energy sources with zero variable cost, such as wind power and solar photovoltaics (PV) – as well as the low cost of natural gas – translates into lower avoided costs for energy efficiency programmes in the electricity sector, potentially diluting the cost-effectiveness of efficiency measures going forward.18 Combined with the variability of rapidly growing solar PV and wind power, the result is greater interest among policy makers in leveraging future energy efficiency measures with advances in demand response, distributed generation and storage, and electric vehicles (EVs), in order to provide various system services that reflect location- and time-specific needs of the power grid.19 ( See Systems Integration chapter.) Nonetheless, as of 2015, incremental energy efficiency measures remained among the least-cost electricity resources in the United States.20

Only 34%

of global energy use falls under the reach of energy efficiency policies and mandates.

Energy efficiency policies come in the form of incentives or outright mandates, such as energy performance standards for appliances and equipment, building energy codes and vehicle fuel economy standards.21 As of 2017, only 34% of global energy use fell under the reach of energy efficiency policies and mandates.22 New policy coverage was attributed largely to equipment stock turnover and to new goods being covered by existing standards, rather than to the adoption of new standards.23


The average efficiency requirements (i.e., stringency) of energy efficiency mandates rose at a faster rate in 2017 than in 2016, but the expansion of policy coverage slowed.24 The increase in the stringency of policy mandates was 0.5% in 2017, slightly below the 0.6% annual average since 2011, and was concentrated mostly in the transport sector on account of rising fuel economy standards.25 ( See Energy Efficiency section in Policy Landscape chapter.)

Global primaryii energy intensity decreased more than 10% during the five-year period between 2012 and 2017, at an average annual rate of 2.2%.26 ( See Figure 55.) The total primary energy supply grew 5.9% over the same period (average annual growth of 1.2%).27 In other words, if energy demand had moved in tandemiii with global economic growth (no reduction in energy intensity), primary energy demand would have risen 18% (3.4% per year).28

Figure 55
Source: Enerdata. See endnote 26 for this chapter.
Figure 56
Source: Enerdata. See endnote 29 for this chapter.

All regions of the world showed some improvement in the energy intensity of their economic activities between 2012 and 2017.29 ( See Figure 56.) Asia (led by China) had the most marked decline in energy intensity during the period – an annual average drop of 3.6% – as the share of energy-intensive industry and commerce continued to shrink relative to all other economic activity, and as manufacturing facilities became more efficient.30 Europe improved at an average annual rate of 2%, followed by North America and Oceania (1.7%); other regions observed only marginal improvements in energy intensity (less than 1% annual average).31

Global primary energy intensity decreased more than


between 2012 and 2017.

Despite the ongoing advances in energy efficiency in many economies and across various end-use sectors, total energy demand continues to rise in regions with rapid economic growth and improved access to energy. In some mature economies, however, growth in total energy demand has long since levelled off and even begun to retract. For example, as of 2017 primary energy demand in the United Kingdom was at its lowest level since 1964, and Germany’s primary energy demand was more than 14% below its historical peak in 1979.32


Collectively, energy demand in member countries of the Organisation for Economic Co-operation and Development (OECD) reached a historical peak in 2007, which also coincided with the onset of a global economic downturn.33 ( See Figure 57.) Despite sustained economic recovery and growth since then, that peak remains unchallenged. Meanwhile, energy demand in non-OECD countries, as a whole, continues to rise. In China, the world’s largest energy consumer, total annual energy demand fell slightly in 2016 – its first decline since 1997 – before reaching a new high in 2017.34


Advances in energy efficiency are most visible in the various end-uses of energy – such as road vehicles, appliances and lighting – and are best examined in the context of final energy use. But gains in energy efficiency also occur when primary energy sources are transformed and converted into various useful secondary forms of energy – such as the production of vehicle fuels in oil refineries and electricity generation in power plants. Those gains are examined in the context of primary energy use.

Figure 57
Source: OECD/IEA. See endnote 33 for this chapter.

In 2016, the global total primary energy supply (TPES) was 576 exajoules (EJ).35 Each year, more than 23% of TPES is dissipated through various transformation processes, the bulk of which is lost during electricity generation.36 The energy industry itself accounts for another 6% of TPES through its net demand for energyiv for purposes such as the operation of oil refineries and the mining and extraction of fossil fuels. Less than 2% of TPES goes to “non-productive” losses, which occur mainly during the transmission and distribution of electricity.37

What remains of TPES – amounting to 400 EJ in 2016 – is the energy available to meet various final energy uses.38 This total final consumption of energy (TFC)v includes all electricity delivered to final customers (nearly 19% of TFC), the final uses of fuels for work and heat (72%), and various non-energy uses (about 9%).39 Broken down by sector, in 2016, the buildings sector consumed 30% of TFC, industry consumed 29%, and transport consumed 29%.40 The remainder of TFC is consumed in other sectors – including agriculture and forestry (2.1% in 2016) – and for non-energy applications (9.1% in 2016), mainly various industrial uses such as feedstocks for petrochemical manufacturing. 41 The relative distribution of final energy demand among the three largest sectors shifted slightly between 2010 and 2016, with the shares of industry and buildings each shedding about one percentage point and the share of transport rising by more than one percentage point; the share of non-energy applications rose somewhat.42

iThere is no single direct measure of economy-wide energy efficiency changes, but energy intensity has stood in as a proxy for aggregate efficiency. Energy intensity is an imperfect substitute because it reflects not only changes in relative energy efficiency but also structural changes in economic activity (such as a shift from heavy industry towards services and commerce) regardless of energy efficiency changes within each economic sector.i

iiThe GSR discusses renewable energy mostly in the context of final energy supply, but primary energy is highly relevant in the context of energy efficiency as it pertains to the conversion and ultimate disposition of primary energy supply, including electricity generation.ii

iiiThis is only to highlight the relative scale of the effect of declining energy intensity on overall energy demand. This does not take into account unknown feedback from higher energy intensity on economic growth. In other words, global economic growth might not have been as great over the observed period if not for the benefit of more efficient use of energy in economic activity.iv

ivThis includes energy input in blast furnaces and coke ovens.iv

vTotal final consumption includes energy demand in all end-use sectors, which include industry, transport, buildings and agriculture, as well as non-energy uses, such as the use of fossil fuel in production of fertiliser. It excludes international marine and aviation bunker fuels, except at the global level, where both are included in the transport sector, from OECD/IEA, World Energy Statistics and Balances, 2018 edition (Paris: 2018). See (Total) Final energy consumption in Glossary for differentiation between TFC and TFEC.v