Author: Andy Hackett, Head of Policy, Centre for Net Zero

Ofgem’s recent review of “typical” household energy use finds that demand has fallen sharply since the 2022 price shock and shows little sign of rebounding. While this decline would traditionally be seen as a success of energy efficiency and climate policy, it also points to a longer trend and deeper problem in Britain’s energy system. Where costs are increasingly fixed, lower consumption does not reduce total costs – it raises the price per unit needed to recover them and suppresses electrification in particular. This blog examines British electricity demand in an international and historical context, and the dynamic known as the “doom loop” which risks locking in higher bills at a time when they urgently need to fall.

Electricity is the future and Britain is behind

The International Energy Agency (IEA) has this year declared the arrival of the “Age of Electricity”, underpinned by strong global demand growth. Global electricity demand is projected to grow around 2.5 times faster than overall energy demand. Electrification of buildings, transport and industry is accelerating, while AI and cooling demand are emerging as major new drivers.[1] When accounting for energy losses, electricity is already the single largest supplier of useful energy.[2]

Source: Nat Bullard (2026), using IIASA, Ember analysis (left); IEA, Electricity 2026 (right)

But the electricity age has not arrived for everyone. Emerging economies account for most of recent demand growth – and roughly 80% of projected increases – with China alone responsible for close to half of the increase to 2030. This is partly a story of convergence from a lower baseline, but increasingly reflects a different development pathway: falling costs of solar, storage and electric technologies are enabling “fast track” electrification. By contrast, fuel-switching away from fossil dependency remains slow in many advanced economies. US electricity demand has been declining or flat from a high baseline, though is expected to see renewed growth from industrial reshoring, electrified manufacturing and data centres. Europe anticipates slower demand growth at around 2% per year, following a period of relatively flat demand.[3] The UK stands out for the steepest and most persistent decline in electricity demand, decreasing by around 20% since 2005, or 30% per capita.

Per capita electricity demand: international comparisons

Source: Our World in Data, using data from Ember (2026)

Part of the decline in demand reflects energy efficiency improvements, but that is hardly a trend unique to Britain. More significant is the pace of deindustrialisation, evident in the fall in overall energy use. Claims that Britain is leading a new “green” industrial revolution, as it did three centuries ago, gloss over the absence of a comparable industrial base today. Deindustrialisation is also only part of the story. Domestic demand remains heavily reliant on gas and oil, with heat electrification among the lowest in Europe. Overall, electricity accounts for only a fifth of today’s final energy consumption.

Total energy use in Great Britain, by sector, since 1970

Source: Centre for Net Zero (2026), using DESNZ DUKES 2025.

​​Britain’s weak electricity demand sits awkwardly alongside its claims of entering a “new era of clean electricity”. Its supply-side achievements are undeniable. For its size, the UK has been a world leader in building out renewables and phasing out coal, with emissions from energy supply falling by around 70% between 1990 and 2020. The most recent renewables auction procured a record near 15 GW of new capacity. However, Britain is racing ahead in building a clean electricity system without building the electric economy to use it, failing to translate this into benefits for consumers.

Low demand, high prices

Weak electricity demand would be less concerning if electricity were cheap and competitive. It is not. Britain has among the highest retail electricity prices in the developed world, particularly for industry. While the 2022 gas crisis amplified the spike, real electricity prices had already been trending upward for much of the previous two decades. In this context, low demand is not simply an outcome of efficiency, but a signal that electricity is failing to compete in key end uses.

UK households face among the highest prices for electricity (IEA energy prices, 2025)

Source: IEA (2026)

What is often missing from discussions of high electricity prices is the interaction between prices, demand and system costs. When prices rise, demand falls. Households substitute towards cheaper alternatives – most notably gas for heating – while higher prices erode real purchasing power. Price elasticity in energy demand is well established: sustained increases dampen consumption, even if the effect varies by sector and over time.[4] As shown below, there appears to be some correlation between the two: consumption peaked in the mid-2000s when real prices were lower, and has weakened as prices have risen since.

Electricity demand vs. Real Electricity Price (% changes relative to 1970)

Source: Centre for Net Zero (2026), using DESNZ DUKES 2025. Table 1.1.5; DESNZ, Historical electricity data (2025), using ONS Retail Price Index (RPI) Table 1.1.5.

This pattern is reflected internationally. The chart below plots the percentage change in electricity prices against changes in electricity demand per capita across countries. Those experiencing falling prices tend to see rising demand, while those with rising prices tend to see stagnant or declining demand. The data are very noisy and the relationship is not necessarily causal, but the direction is intuitive: where electricity becomes relatively more expensive over time, demand growth is typically weaker.

International comparisons: change in electricity prices vs. change in electricity demand per capita (selected countries with available data)

Source: Centre for Net Zero (2026), using IEA End-Use Energy Prices (residential electricity, USD/MWh). Ember Global Electricity Review (demand per capita, MWh).

The above chart combines very different economies, and particularly those starting from very low baselines have experienced huge electricity growth even when prices rise. Restricting the sample to OECD countries – broadly comparable in income levels, institutions and industrial structure – suggests that among advanced economies, the UK still stands out for the combination of relatively large price increases and weak or negative growth in electricity demand per capita.

OECD comparisons: change in electricity prices vs. change in electricity demand per capita

Source: Centre for Net Zero (2026), using IEA End-Use Energy Prices (residential electricity, USD/MWh). Ember Global Electricity Review (demand per capita, MWh).

Energy transitions and price elasticities

The literature on long-run elasticities of energy demand – notably, Roger Fouquet’s historical analysis – shows that in early phases of electrification, demand for energy services is elastic: as incomes rise and service costs fall, consumption expands rapidly.[5] Over time, elasticities decline as saturation effects set in. Historically, new energy systems expanded because the effective cost of useful energy services fell through efficiency gains and innovation.

Coal, gas and electricity spread because they became cheaper, cleaner, more convenient or more productive. Falling costs stimulated demand; higher volumes justified infrastructure investment; and scale reduced costs further. Successful transitions combined falling service costs with infrastructure expansion and coordinated end-use conversion. The UK’s conversion from town gas to natural gas in the late 1960s and 1970s is instructive: over little more than a decade, 35 million appliances were replaced across 13 million homes.[6] Infrastructure and demand-side change moved together, enabling a rapid system-wide shift.

Over the long run, advanced economies tend to devote a relatively stable share of income to energy, often around 7-8% of GDP.[7] Higher prices tend to drive energy productivity – through efficiency, innovation and structural change – reducing energy intensity over time. But this “self-correction” applies to total energy use, not individual fuels. Electricity competes with close substitutes, such as gas for heating. If it remains persistently more expensive, households and firms will not switch, even as overall energy spending stabilises. The result is a system where total demand adjusts, but electricity remains structurally under-consumed. Decarbonisation depends not just on how much energy is used, but on which fuels are used.

Energy prices versus energy productivity (average values for 1970 to 2019)

Source: Bashmakov et al. (2024)

Efficiency is a good thing, but in the UK electricity is losing out to other energy carriers. It is not becoming the cheapest or most convenient carrier in key end uses – particularly in heating, where gas remains structurally advantaged on price. Under these conditions, price elasticity does not just slow demand growth; it diverts consumption towards incumbent fuels. Rather than reinforcing electrification, relative prices are sustaining fossil fuel demand. That is a policy choice, shaped by how the costs of the energy system are recovered and distributed.

The UK electricity “doom loop”

Decarbonising electricity systems are capital-intensive networks characterised by high fixed costs and relatively low marginal costs. A large share of UK electricity bills reflects network charges, policy levies and capacity mechanisms rather than wholesale energy costs, and these costs are largely recovered on a per-unit basis. When demand stagnates, those costs must be spread across fewer kilowatt-hours. Average costs rise, pushing up unit prices and further discouraging electrification, entrenching higher costs. This creates the mutually reinforcing dynamic known as the “doom loop”, closely related to the “utility death spiral” in which declining consumption raises unit prices and further accelerates demand reduction or grid defection.[8]

The problem is compounded by how the UK structures energy prices. Electricity carries a disproportionately high share of policy and network costs, and is taxed far more heavily than gas. The result is not just a system that amplifies weak demand, but one that actively reinforces substitution away from electricity in key end uses such as heating.

Within Europe, the UK has among the highest electricity bills, and one of the highest share of policy and networks costs

Source: IEA (2026)

Grid investment is also being planned against ambitious net zero scenarios rather than probabilistic demand trajectories. This means operators are building for a highly electrified future economy that has not yet materialised. This is not inherently wrong – infrastructure must anticipate demand – but it becomes destabilising when the cost recovery model assumes rapid volume growth that does not arrive. The mismatch is becoming a political issue in maintaining support for clean power.

Final electricity demand to 2024 with NESO projection to 2050

Source: Centre for Net Zero (2026), using DESNZ DUKES 2025. Table 1.1.5; NESO Future Energy Scenarios (2025).

The “doom loop” has distributional consequences. Lower-income households already devote a larger share of expenditure to energy. If higher-income households electrify selectively, self-generate, adopt behind-the-meter solutions or partially defect from the grid, the residual cost burden falls disproportionately on those least able to adjust. This risks locking in a system where the fixed costs of decarbonisation are borne by a shrinking and increasingly price-sensitive demand base. The result is economically regressive and exacerbates the political challenges of sustaining support for net zero.

Household bills are becoming more and more dominant for lower income groups

Source: Resolution foundation (2026), using ONS, Living Costs and Food Survey.

Data centres are not a substitute for economy-wide electrification

Putting aside wider debates about AI, future electricity demand from data centres is seen by many as a looming threat to the grid. Headlines warn they could add 50 GW of new capacity in Britain – roughly doubling today’s peak demand – based on the projects in the demand connections queue. To others, they are precisely the kind of modern industrial load the UK needs: a source of new electricity demand that could help break the demand “doom loop”. The reality lies somewhere between the two.

There is no doubt that data centres represent a significant prospective load. Globally, however, they account for a relatively modest share of projected electricity demand growth this decade – around 8% between 2024 and 2030 in IEA-based projections. They matter, but they are not the dominant structural driver of electrification.

Global electricity demand growth and projections, by sector and end-use, 2015-2030

Source: IEA (2026)

Clearly, adding up prospective projects in the British demand connections queue is not a projection of future load. When the generation connection queue in Britain exceeded 800 GW – more than four times the capacity required to deliver clean power by 2030 – no one seriously thought that every megawatt in that queue would be built. The demand queue is similar, with developers securing land and grid position well before a hyperscaler tenant is contracted, and many projects will never materialise at scale.

For all the rhetoric around AI expansion, rapid build-out of data centres will face limits. In the US, the majority of proposed data centres remain stuck in planning (see below). Hyperscale operators prioritise cost, reliability and speed of connection. On each of these metrics, Britain faces particular structural constraints: high wholesale prices, rising network charges and extended grid connection queues. In response, operators are increasingly exploring private power purchase agreements, co-location with generation, or behind-the-meter solutions. If large loads bypass the public network, they do little to improve utilisation of sunk system assets – and therefore little to resolve the average-cost pressures at the heart of the “doom loop”.

Planned data centres in the US are largely stuck in planning

Source: Nat Bullard (2026)

The UK’s strategy should not fear data centres as a source of new electricity demand, but focus on integrating it in a way that supports the grid – locating near abundant renewables, non-firm grid connections, and their own dedicated clean power and storage – and ensure that developers pay their way. However, this is neither guaranteed nor sufficient, especially in the short term. Broad-based electrification of heat, transport and industry is the only reliable foundation for sustained, economy-wide electricity demand growth.

Build demand or pay the price

The task is not to abandon clean power, but to couple it with rapid electrification – accelerating the conditions under which its benefits reach consumers. Whitehall’s response to the current energy challenge has, until recently, leaned heavily on familiar supply-side measures, bringing forward auctions for more renewables.

Recent announcements signal an encouraging shift in response to the current energy crisis. Increased support to switch from oil heating, a bit more investment in solar and storage for social housing, and reforms to ease the installation of low-carbon technologies are all welcome measures.

But they remain piecemeal. They do not address the structural features of the system that continue to suppress electricity demand: the relative price disadvantage of electricity compared to gas, and the recovery of growing fixed system costs from bills. Without reform here, the risk is that incremental demand-side support is offset by persistently weak underlying incentives.

The government should go further. Demand must be treated not as a by-product of clean supply, but as a central pillar of energy strategy – supported through pricing reform, targeted subsidy and coordinated deployment at scale. Without this, Britain risks building a clean electricity system that remains underused, expensive, and politically fragile.

---------------------------------------------------------------------------------------------------------------------------------------

[1] IEA (2026) Electricity 2026: Analysis and forecast to 2030. Paris: International Energy Agency

[2] Ember (2025), The long march of electrification

[3] IEA (2026); Ember (2025), Powering China’s New Era of Green Electrification

[4] Castle, J.L., Hendry, D.F. and Martinez, A.B. (2023) ‘The historical role of energy in UK inflation and productivity with implications for price inflation’, Energy Economics, 126, 106947

[5] Fouquet, R. (2014) ‘Long-run demand for energy services: Income and price elasticities over two hundred years’, Review of Environmental Economics and Policy, 8(2), pp. 186–207.; Fouquet, R. (2016) ‘Path dependence in energy systems and economic development’, Energy Policy, 88, pp. 2–8.

[6] Stern, J. (1987), The Economics of Natural Gas. London: Macmillan; Williams, T.I. (1981), A History of the British Gas Industry. Oxford: Oxford University Press.

[7] Bashmakov, I., Grubb, M., Drummond, P., Lowe, R., Myshak, A. and Hinder, B. (2024) ‘“Minus 1” and energy costs constants: Empirical evidence, theory and policy implications’, Structural Change and Economic Dynamics, 71, pp. 95–115.

[8] Felder and Athawale (2014) ‘The life and death of the utility death spiral’, The Electricity Journal, 27(6), pp. 9–17.