The effects of public campaign funding: Evidence from Seattle’s Democracy Voucher program

Do Democracy Vouchers help democracy?

“…Both this paper and Griffith and Noonen (2022) find increases in small donations, but including 2021 data reduces the magnitude substantially: I estimate that 95%-156% more dollars were given in small donations to city council candidates, compared with their estimate of 350%. While Griffith and Noonen (2022) estimate a 53% increase in total contributions per city council race, I estimate a 46% decrease in total contributions to city council candidates. Including data from an additional election is useful in determining whether patterns from the first two elections with Democracy Vouchers are persistent over time.…”

“…Another option would be to perform the analysis at the race level. Race-level analysis was the focus of Griffith and Noonen (2022), and I complement their analysis by studying outcomes at the city level.…”

“…One advantage of my approach is that I am able to provide evidence of parallel pre‐trends, while Griffith and Noonen (2022) do not have parallel pre‐trends for their campaign finance outcomes. Both this paper and Griffith and Noonen (2022) find increases in small donations, but including 2021 data reduces the magnitude substantially: I estimate that 95%–156% more dollars were given in small donations to city council candidates, compared with their estimate of 350%. While Griffith and Noonen (2022) estimate a 53% increase in total contributions per city council race, I estimate a 46% decrease in total contributions to city council candidates.…”

“…By contrast, Yorgason (2021) finds that voucher availability made the donor pool less representative of the electorate. The most closely related paper is Griffith and Noonen (2022), which uses difference-in-differences to study Democracy Vouchers' effects on campaign finance outcomes. I perform my campaign finance analysis using the Ferman and Pinto (2019) method and my control group consists of other cities in King County, while their control group is composed of several large cities in Washington and California.…”

“…I also include data from Seattle's 2021 election, while their sample ends in 2020. One advantage of my approach is that I am able to provide evidence of parallel pre-trends, while Griffith and Noonen (2022) do not have parallel pre-trends for their campaign finance outcomes. Both this paper and Griffith and Noonen (2022) find increases in small donations, but including 2021 data reduces the magnitude substantially: I estimate that 95%-156% more dollars were given in small donations to city council candidates, compared with their estimate of 350%.…”