How does political geography impede efforts to mitigate climate change? Climate policies impose costs on geographically concentrated carbon-intensive industries and their surrounding communities. This is politically costly: evidence demonstrates voters punish pro-climate parties for job losses in fossil fuel-producing regions. Existing studies often focus, however, on ex-post electoral consequences. We know little about how voters in these communities respond ex-ante to proposed climate polices. I study this question in the context of the 2019 Australian federal election. Australia is a top fossil producer, and carbon-intensive activity is highly concentrated in specific regions. The 2019 election was dubbed the 'climate election' because climate split the major parties: the Labor Party proposed a package of reforms, while the Liberal-National Coalition advocated for fossil fuel expansion. I assemble a novel database of vote shares in over 4,000 polling places over seven elections. I use difference-in-differences to estimate whether a community's fossil fuel intensity caused it to punish the pro-climate party in 2019, compared to a matched control group. To unpack whether perceptions of material cost drive election results, I compare communities facing short and long-term policy risks. I expect the pro-climate party to suffer more in communities dominated by coal-fired power generation and threatened by renewable energy expansion, than those surrounding metallurgical coal mining for export. My findings shed light on the electoral consequences of geographically concentrated 'policy losers' in majoritarian systems. I help to reveal the mechanisms driving these dynamics and provide insights for policymakers developing political strategies for climate change mitigation.