Batteries have come down in prices faster than anticipated
Battery costs have collapsed 99% since 1991 — but the dynamic driving that fall is about to matter far more than the number itself.
Something unusual happened on Australia's east coast electricity grid last summer. On the hottest days, when gas-fired power stations were supposed to be indispensable, they barely ran. Gas generation fell to its lowest share in 25 years. Batteries, quietly and without much ceremony, had taken over the peak-hour job. Wholesale electricity prices are now 47% lower than a year ago. This is not a policy announcement. It is the market telling us something has structurally changed.
Wright's Law explains why battery costs keep falling — and why gas costs don't
The mechanism behind that cost collapse is one of the more elegant dynamics in industrial economics. It is called Wright's Law, named for an aeronautical engineer who noticed in the 1930s that every time aeroplane production doubled, the labour cost per unit fell by a predictable percentage. The same pattern holds for solar panels, semiconductors, and now batteries. CSIRO research cited in the latest GenCost report confirms that every time global battery storage capacity doubles, the cost of manufacturing battery cells falls by around 21%. Between 2024 and 2025 alone, global battery storage capacity jumped 40%. Do that arithmetic and the price trajectory becomes less surprising.
What makes this dynamic particularly potent is that it feeds on itself. Cheaper batteries create new uses. New uses create more demand. More demand drives more manufacturing. More manufacturing drives the next round of cost reductions. Government policy did not invent this cycle, but it has been accelerating it. Investment incentives and renewable energy targets, including Australia's goal of 82% renewable generation by 2030, pull more battery projects into existence sooner than the market alone would have. In a conventional demand-stimulus situation, you would worry that subsidising buyers simply inflates prices, with the grant getting capitalised into the asset cost and the seller capturing the benefit rather than the buyer. Battery markets are different. Supply can expand, and the act of expanding it makes the next unit cheaper. The subsidy does not just shift demand along an existing supply curve; it shifts the curve itself.
The subsidy does not just shift demand along an existing supply curve; it shifts the curve itself.
Gas turbines face the opposite dynamic: scarce supply, rising costs, no learning curve
This stands in instructive contrast to what is happening on the gas side of the ledger. Gas turbines are produced in small volumes by a handful of global manufacturers. They are complex machines, the supply chain is narrow, and demand was low for most of the past decade. Now, partly due to data centre power requirements in the United States, demand for turbines has surged. Delivery times have blown out to four years. None of the dynamics that have driven battery costs down apply here: there is no Wright's Law effect when production volumes are low and oligopolistic, no learning curve steep enough to offset the structural scarcity. Australia's newest gas peaker, Kurri Kurri in the Hunter Valley, can only run for about ten hours before it needs more than a day to refuel. A modern eight-hour grid battery, by comparison, recharges overnight and is ready again by morning.
The real question is not gas versus batteries — it is how much backup the grid actually needs
This does not mean gas exits the grid tomorrow. Battery storage is increasingly dominant at peak times, but a grid running on wind and solar will occasionally face extended periods of low generation, cloudy and still stretches that coincide with high demand. Some form of dispatchable backup remains necessary. The more honest version of the energy debate is not gas versus batteries but rather how much stored fuel backup capacity the grid genuinely needs, and whether it can run on something other than fossil gas. The infrastructure question matters because new gas pipelines and gas fields carry a real stranded-asset risk as global carbon constraints tighten over the next two decades.
Demand-side subsidies work differently when supply has a steep learning curve
There is a broader point here about policy and technology markets that gets underappreciated in the debate about government incentives. The standard critique of demand-side subsidies, one The Bearing has applied to housing policy before, is that they inflate prices when supply is constrained. That critique is well-founded in markets where supply genuinely cannot respond. But technology markets with steep learning curves operate differently. In those markets, policy that accelerates adoption can genuinely shift the cost structure downward for everyone, not just the subsidy recipients. The subsidy does not disappear into seller margins; it funds the production volume that makes the next generation of the technology cheaper for buyers who never received a cent of government support.
Wright's Law does not care about ideology. It cares about cumulative production. And right now, cumulative battery production is compounding faster than almost anyone forecast.
Sources
The Conversation — Batteries are getting cheaper while gas gets pricier. Here's why
The Bearing — Budget's Innovation Bet: Who Bears the Cost When Incentives Misfire?
The Bearing — Gas tax is all hot air
Frequently Asked Questions
Why have battery storage costs fallen so much?
Battery costs follow Wright's Law: every time global production capacity doubles, manufacturing costs fall by around 21%. Global battery storage capacity grew 40% between 2024 and 2025 alone, compressing that learning curve faster than most forecasters expected. The result is a cost fall of more than 99% since 1991.
Do government subsidies for batteries just inflate prices like housing grants do?
Not in the same way. Housing subsidies inflate prices when supply cannot respond — the grant gets absorbed by sellers. Battery markets operate on a steep learning curve where expanding supply actually reduces the cost of the next unit. A subsidy that pulls forward production volume ends up making batteries cheaper for everyone, including future buyers who received no subsidy at all.
Can batteries fully replace gas on the Australian grid?
Not yet, and possibly not on their own. A grid powered by wind and solar will face extended low-generation periods — still, cloudy days — where batteries alone cannot bridge the gap. Some dispatchable backup remains necessary.
What is the risk of investing in new gas infrastructure in Australia?
New gas pipelines and gas fields have long asset lives, typically measured in decades, during which global carbon constraints are likely to tighten significantly. If demand for gas falls faster than current projections — as battery adoption has repeatedly outpaced forecasts — those assets could be written down well before their costs are recovered, a stranded-asset outcome investors in coal have already experienced.
Why is gas getting more expensive while batteries get cheaper?
Gas turbines are complex machines produced in low volumes by a small number of manufacturers, so there is no meaningful learning curve driving costs down. Demand has surged recently, partly because US data centres require large amounts of reliable power, pushing delivery times to four years. Unlike batteries, gas turbines face the scarcity economics of a narrow-supply, high-demand market with no self-reinforcing cost-reduction cycle.