Words matter, particularly with poll questions
2 polls on the same issue a month apart come to different conclusions
Due to a reporting error, an earlier version of this story gave the incorrect date for the initial poll on the so-called millionaire tax. The two polls were a year apart, not a month apart.
A POLL released in mid-January 2021 indicated 69 percent of registered voters in Massachusetts either strongly support or somewhat support the constitutional amendment poised to appear on the November ballot creating a higher tax on income over $1 million.
On Wednesday, a poll of likely voters came to a very different conclusion on the proposed constitutional amendment — only 37.2 percent of those surveyed strongly supported or somewhat supported the higher tax.
Nothing much changed over that year-long period between the polls except who was asking the question, who was being asked the question, and the way in which the question was asked.
The poll said the ballot question would create an additional 4 percent tax on income over $1 million a year, a trigger that would rise with inflation.
The poll in February 2022 was sponsored by the right-leaning Fiscal Alliance Foundation and conducted by Jim Eltringham of Advantage Inc., a Republican-affiliated polling company in Washington, DC. The February poll said legislative leaders have placed a referendum on the November ballot that would amend the state constitution to allow lawmakers to raise the income tax on some high-income earners and middle-class small businesses.
The questions on both polls were accurate, but they emphasized different things, which probably influenced the poll results.
The January 2021 poll question specified how much the tax would be (4 percent) and how it would apply only to income over $1 million. The February 2022 question never mentioned the figure of $1 million. Instead, it placed heavy emphasis on the fact that legislative leaders crafted the question, which would allow them to raise the income tax on “some high-income earners and middle-class small businesses.”
The January 2021 question overall is more precise about how the tax would work and makes clear that it would only apply to millionaires. The February 2022 question is more vague about the particulars of the tax, but very specific about who put it forward and, to some degree, who might be affected.
Eltringham said he wasn’t trying to include messaging in the poll question to steer respondents to a desired conclusion. “We really tried to strip down and make it as bland as possible,” he said. “Again, there’s going to be a campaign around this so there’s going to be messaging around it. It’s always going to be in some kind of context. We really don’t want to use terms that would try to skew it one way or another.”
Eltringham said another question in the February poll did have “some messaging.” The question said that if the ballot measure passes, the Legislature would be able to raise the income tax from 5 percent to 9 percent for some high-income earners and middle-class small businesses — “meaning that those people could see their taxes increase by 80 percent.” The question asked if the increase was too high, too low, or just about right.
Nearly 70 percent of those surveyed said the increase was too high, nearly 2 percent said it was too low, and 18.3 percent said it was just about right, according to the poll
Eltringham said he did not want to refer to the million-dollar threshold — as the January poll did — because it would sway those answering the poll questions. “There is a fair amount of connotation when you start talking about it as a millionaire’s tax, and you start conjuring up images of people in top hats and monocles,” he said. “So we didn’t want to talk about it in terms of that.”
As for the inaccurate claim in the question that those affected would see their taxes increase 80 percent, Eltringham said the 80 percent increase in the tax rate was “an important talking point.”
The bottom line is words matter. Trying to boil down a complicated ballot question into a few sentences isn’t easy and involves a series of value judgments about what’s important and what’s not. Those value judgments can sway how voters view an issue and ultimately how they vote on it, which is why the actual ballot question summary and the “yes” argument in a voter information guide are currently being challenged in court.The lawsuit questions the accuracy of the claim that the money raised by the tax can only be used to fund new spending on education and transportation, when in fact the money is fungible and subject to legislative appropriation.
Paul Craney of the Massachusetts Fiscal Alliance, who helped present the results of the poll on the ballot question Wednesday, is also part of the group challenging the wording of the information voters will see in November.