The standard story goes like this: wind and solar have near-zero marginal costs. As their share in the grid grows, they displace expensive gas peakers, drive down wholesale electricity prices, and ultimately make energy cheaper for everyone — including the poorest households. The story is true. It is also incomplete.
A Fixed Effects panel regression across 22 EU member states from 2010 to 2020 tells a more complicated story. The relationship between renewable energy share and an aggregate energy poverty index is not monotonic. It is nonlinear. In the early stages of the transition, increasing renewable share is associated with higher energy poverty — before the relationship reverses.
The mechanism: transition costs arrive before transition benefits
The transition introduces costs before it delivers savings. Grid modernisation, storage infrastructure, backup capacity for intermittent sources, and the interconnection investments needed to balance variable supply all appear as levies on electricity bills years before the market-clearing effect of cheap renewables fully materialises. For a low-income household, the cost arrives in the next quarterly bill. The benefit arrives — if it arrives — on a decade-long horizon.
The regression isolates this dynamic. Holding constant GDP per capita, heating degree days (a proxy for climate), housing quality, and country fixed effects, an increase in renewable share from 10% to 25% is associated with an increase in the energy poverty index of approximately 3.2 points. At shares above roughly 35–40%, the direction reverses and the relationship becomes negative — consistent with the long-run theory.
Stylised relationship: RES share vs energy poverty index (22 EU countries, 2010–2020)
The efficiency breakpoint
A separate finding from the regression concerns the interaction between renewable share and housing energy efficiency. The protective effect of efficient housing — measured by the EU's energy performance certificate (EPC) rating distribution — becomes statistically significant only at an aggregate efficiency index above approximately 67.5 on a 0–100 scale. Below that threshold, improving the renewable mix provides no statistically detectable poverty-reduction benefit.
This suggests a sequencing constraint: policy that expands renewables without first improving housing stock for the most vulnerable households may be spending money in the wrong order. Efficient housing reduces the quantity of energy a household needs to purchase; renewable expansion reduces the price per unit. The two interventions are complements, but only the first is sufficient on its own.
Poverty index reduction by housing efficiency group — estimated marginal effects
Policy implications
The finding does not argue against the energy transition. Countries with high renewable shares — Denmark, Austria, Sweden — do have lower energy poverty rates than coal-dependent economies at comparable income levels. The long-run trajectory is favourable.
It does argue for transition management. The households most exposed to the upfront cost of transition — low-income renters in poorly insulated buildings in grid-expansion zones — are the least able to absorb it. Without targeted protection during the transition phase, the aggregate poverty index can worsen even as aggregate emission targets improve. A transition that succeeds on climate metrics while failing on equity metrics is not a just transition.