Authors: Chloe Lucia Nemo Ramirez, Guest Contributor; Joe Grainger, Trials and Analysis Project Manager; Andrew Schein, Director of Trials and Analysis.

This blog explores how non-energy costs on electricity bills affect heat pump running costs. You can experiment with the impact yourself using our Heat Pump Running Costs Simulator.

An engineering marvel

Heat pumps are remarkably efficient machines. Because they move heat rather than generate it, they deliver roughly three units of heat for every unit of electricity they consume. A gas boiler, by contrast, converts fuel to heat at something like 90% efficiency, meaning you burn a little more than you get. On pure physics, the heat pump wins easily.

So why, when many households in Great Britain switch from a gas boiler to a heat pump, do their energy bills barely change? Our recent research paper, now in revise and resubmit at the American Economic Journal: Economic Policy, provides the most rigorous answer yet. Using data from over a thousand heat pump installations by Octopus Energy, we find that the average household reduces gas consumption by around 8,700 kWh per year and increases electricity consumption by around 2,500 kWh. The implied efficiency is right in line with what engineers would predict.

The efficiency is real. The bill savings aren't guaranteed, at least not yet. The reason is straightforward: electricity in Great Britain costs three to four times as much per kWh as gas. Under current standard tariffs, we estimate that the average heat pump household pays about the same as a comparable gas-heated home; or slightly less, if they disconnect from the gas grid and stop paying the standing charge. A notable exception is if the household is on a time-of-use tariff, which we found saved around 18% on electricity costs – enough to make a heat pump cheaper than a boiler even with unit rates stacked against it.

This fact is well-known in energy circles. A technology that uses dramatically less total energy, and produces dramatically fewer carbon emissions, cannot guarantee private financial reward for the household that adopts it. That seems like a policy problem worth understanding.

The gap between private prices and social costs

The economists Severin Borenstein and James Bushnell have written two papers that, taken together, provide a clean diagnosis of what's going on. Their core argument is that retail energy prices in regulated markets are distorted in ways that systematically disadvantage electricity relative to gas.

The first distortion is energy policy costs loaded onto electricity bills, mostly via unit rates. These levies are how the country has funded its renewables buildout, taking the overall share of low carbon generation to roughly three-quarters of Great Britain’s electricity today. That buildout is an achievement in terms of emissions reductions, but funding them through the electricity unit rate means that every additional kWh of electricity consumed bears a cost that has nothing to do with the marginal cost of producing it. That is a policy choice that has created an avoidable market distortion.

The second distortion runs in the opposite direction on gas. Household gas consumption is currently exempt from the UK Emissions Trading Scheme, meaning it faces no carbon price at the point of use. Yet, gas generates carbon emissions, and those emissions impose costs on society. By excluding these costs from the price households pay, we are underpricing gas relative to its true social cost. On top of that, and even more problematically, the gas used for electricity is covered in the Emissions Trading Scheme. The result is inconsistent policies that lead households to see gas cheaper than it really is and electricity as more expensive.

The third distortion is less straightforward. Electricity infrastructure is highly capital-intensive, and regulators bundle a large share of fixed network costs into per-unit (volumetric) charges. This pushes the retail electricity price above the marginal cost of generating and delivering an additional kWh in most hours – despite the fact that it's peak demand, not average demand, that ultimately determines how much network capacity needs to exist. Gas network costs are also partly recovered this way, but the effect is less pronounced.

Put these together and you get what our paper, following Borenstein and Bushnell, calls "headwinds" for heat pump adoption. Electricity prices are inflated above their marginal cost by fixed cost recovery and policy levies. Gas prices are deflated below their marginal social cost by the absence of carbon pricing. The heat pump sits in the middle, its real efficiency advantage obscured by price signals that point in the wrong direction.

Getting prices “right”: what the numbers say

In a new section of our paper, we combine our empirical estimates of how heat pumps change household energy use with a series of stylised scenarios that progressively remove these distortions and move retail prices closer to marginal social cost. In doing so, we assume that the real-world consumption patterns observed in our study continue to apply and abstract from additional flexibility or behavioural responses to changing prices. The results are striking.

Under then-current prices, our evidence suggests the average heat pump household pays about £10 less per year than a gas boiler household (assuming gas disconnection). If the government's October 2025 policy changes, removing 75% of Renewables Obligation costs and ending the Energy Company Obligation, are reflected in bills, that saving increases to around £63. Removing all explicit electricity policy costs from unit rates pushes it to £106.

Remove electricity network charges from unit rates as well, and the saving rises to £227. Add carbon pricing on household gas at the current UK ETS price, and it reaches £326. Price both fuels at the UK Government's central social cost of carbon, and the heat pump saves roughly £600 per year compared to a gas boiler.

Addressing the electricity-to-gas ratio is Econ 101

The case for rebalancing gas and electricity prices tends to get framed as a green policy priority, something for energy geeks and climate advocates to argue about. We think that framing undersells it.

The underlying principle is basic welfare economics. Prices should reflect marginal social costs. Where they don't, private decisions diverge from socially efficient ones: too much of the expensive-to-society good gets consumed, and too little of the cheap-to-society good. In this case, gas is underpriced relative to its social cost (because its carbon externality goes unpriced), and electricity is overpriced relative to its marginal cost (because fixed costs and policy levies are bundled into unit rates). The result is a private incentive structure that points away from electrification, even when it’s the right call for society.

Each distortion is a choice about how we price energy, and the choices are not equally justified. Taking the policy levies off electricity bills is about as clear-cut as energy policy gets, since they have nothing to do with the cost of a kWh and could be funded another way. Carbon pricing on gas is a no-brainer for emissions reductions. Network cost recovery is a genuinely hard design question, because the cost is unavoidable even though charging for it per kWh is not. The headwinds against electrification, in other words, are decisions, and decisions can be revisited.

This doesn't mean the policy path is simple. Shifting billions of pounds of levy costs off electricity bills means finding another way to fund them – such as through general taxation or standing charges. Carbon pricing on gas is politically contentious and practically complex. Network cost recovery is a genuinely hard regulatory design question.

Still, the economic logic is clear. Getting prices right is not a special favour to heat pumps. It's what you'd want to do anyway, in a world where you cared about efficient resource allocation.

The bigger prize: shrinking network costs, not just reslicing them

Everything above is really about reallocating a fixed pot of network expenditure – levies versus standing charges versus general taxation. That's a legitimate debate, but it doesn't change how much the network costs to build and run. There's a separate, bigger lever: because capacity is sized to meet peak demand, reducing the peak can avoid some of that cost ever being incurred, rather than simply moving it from one bill to another.

Demand flexibility is the mechanism. If heat pumps shift consumption away from the hours when the grid is most strained, less capacity needs to be built or reinforced in the first place. Our paper provides direct evidence on this: a heat-pump-specific time-of-use tariff reduced evening peak demand by 47%. That demand flexibility is the ticket to a genuine overall saving.

Try it yourself

Of course, levy reform involves real trade-offs between efficiency, fairness and political feasibility. Reasonable people can disagree about where those trade-offs should land. We have developed a simulator that lets you experiment yourself to understand policies’ impacts on running costs. The results make clear that the future of heat pumps is not just a question of technology. It is also a question of how we choose to price energy.