We do take into account the falling price of solar panels, but the TVA region is too small to drive the type of learning-by-doing that can reduce panel costs world wide. As panel prices fall, more households from every income group install. Obviously, there is a larger impact on higher income hh's.

The solution isn't obvious. Two-part tariffs or fixed grid access fees could eliminate cost-shifting. But policymakers must grapple with who pays for grid infrastructure in an era of distributed solar. Clean energy transitions require equitable design—this paper shows why.

BUT: the death spiral has a silver lining in most areas. Higher rates incentivize MORE low-income solar adoption. If LMI adoption were 6.3pp higher, the cost shifts would reverse & lower LMI bills. The question: Can we make solar accessible to those who need it most?

We use agent-based modeling + choice experiments (2,307 TVA customers) to simulate this "utility death spiral." The result? $7.8M in additional annual electricity bills ($9.86/customer) for low-income non-adopters by 2051. That's equity destruction disguised as clean energy.

Here's the problem: utilities recover fixed grid costs through per-unit electricity rates. When wealthy customers install solar & reduce their consumption, utilities spread those fixed costs across fewer customers—disproportionately burdening low-income households who can't afford solar.

Find the full paper here. www.sciencedirect.com/science/arti...

We show reciprocity: Connected city leaders who achieve stronger emissions reductions receive more promotions & fiscal transfers—but only in high-target provinces. Suggests patronage ties sustain cooperation through mutual benefit recognition.

Firms in connected cities reduce SO2 emissions ~10% more than those in unconnected cities post-2006. Effect concentrates in private/foreign firms in Eastern China & increases with provincial abatement targets. Reductions driven by lower emissions intensity, not firm exit.

We use 245k firm-year observations from China's environmental database (covers ~85% of pollution), linked to official biographies & administrative data on 325 cities. We also validate using satellite SO2 concentrations and real-time monitored firms.

We study China's 11th 5 year plan (2006-2011), which assigned SO2 targets to provinces but not cities. We measure patronage using past promotions. Using relative tenure as an IV, we identify causal effects by exploiting frequent leader reshuffles & variation in provincial abatement targets.