The state of Washington saw a “blue wave” in the November 2018 midterm elections. Democratic Senator Maria Cantwell was reelected with 58.4% of the vote. In the House of Representative elections, Democrats snatched the 8th Congressional district from Republicans.
And yet, the carbon tax initiative (I-1631) lost, securing only 43.5% of the vote. A number of I-1631 supporters and media outlets blamed “fossil fuel” money for the defeat. The No on I-1631 campaign spent $31.6 million, while the Yes on I-631 campaign spent $16.4 million (Ballotpedia 2018).
But did the I-1631 lose because the opposition spent twice as much money? In this study we test whether the variation in numbers
of television ads across Washington’s four media markets influenced voting behavior. Our analysis fails to support the claim that opposition ads defeated the 2018 carbon tax initiative. We test two hypotheses with voting data aggregated at the Census tract level, tract-level
demographic covariates, and TV ad counts that vary at the media market level. First, all else equal, the lower number of TVadvertisements put out by Yes on I-1631 in relation to No on I-
1631 (henceforth, the ad ratio) is associated with lower electoral support for I-1631. Our second hypothesis pertains to the change in support for I-1631 in relation to another carbon tax initiative, I-732, which was also defeated by Washington voters in 2016 with 40.8% support. While the fossil fuel industry financed the campaign against I-732 as well, it spent only $1.4 million in 2016, a fraction of the $31.6 million spent in 2018 (Ballotpedia 2016). Therefore,we hypothesize that a smaller the Yes/No ad ratio in 2018 will be associated with less support for I-1631 in relation to I-1732.
We do not find statistical evidence of an association between the ad ratio and I-1631 vote share, even after controlling for Census tract-level demographic characteristics. Thus, our analysis does not support the claim that fossil fuel-funded TV advertising influenced the
electoral outcome. Regarding our second hypothesis, we find that the ad ratio is negatively associated with the change in carbon tax support from 2016 to 2018. In other words, the vote share in favor of I-1631 increased with respect to I-732 in areas that received a higher number
of No on I-1631 advertisements relative to Yes on I-1631 advertisements.