Vanessa Brown Calder of the Cato Institute has a response to a recent criticism I made of her and one of her colleagues, Chris Edwards.

Cato had commissioned a poll showing that most people, especially conservatives, look less favorably on liberal proposals to expand access to paid family leave when they are told about some of the costs of those proposals. I thought it was wrong of Calder and Edwards to attempt to use this poll to discredit the very different proposals that some conservatives have made, which do not involve the same costs. For example, the poll tests how people feel about family leave proposals that raise payroll taxes; Rep. Ann Wagner and Senators Joni Ernst, Mike Lee, and Marco Rubio have been talking about proposals that do not raise payroll taxes.

Calder defends this sleight-of-hand by claiming that she was responding only to Rick Santorum, who wasn’t specific about what policy he had in mind. Further, she argues, Santorum urged Republicans to work with Democrats, and Democrats are not interested in the conservative legislators’ ideas. Rubio’s idea is “not popular,” she says, apparently using that phrase to mean that it has no co-sponsors.

If Edwards and Calder had written that conservatives don’t like tax increases, that Cato had decided to spend its donors’ money proving the point, and that any legislator attempting to work on this subject should keep that in mind, I would have had no objection to it. I would have noted, however, that Republicans should not let this polling data keep them from promoting better ideas about paid leave.

But they didn’t write that. Edwards used the poll to attack an idea that it didn’t test: a proposal from the conservative Independent Women’s Forum. Calder chose to end her article by specifically calling out Senators Ernst, Rubio, and Lee, without noting that none of them had endorsed the ideas that the poll tested and all of them had endorsed different ideas that were not tested.

This is sloppy, at best. It’s not a hanging offense, but it’s not truly defensible either, as Calder’s post inadvertently demonstrates.