Complaints about the lack of transparency in how the New York Times compiles its bestseller lists are nothing new. For years, some publishers have complained that the paper’s methodology—relying on point-of-sale data from a secret and select list of stores—can produce skewed and inaccurate results. For publishers of conservative books, the Times‘ approach has long been a headache. Now, one of these publishers has taken action.

Regnery Publishing president and publisher Marji Ross said that, effective immediately, the publisher is cutting all ties with the paper and will no longer use the Times as a basis for providing bonuses to its authors and employees. In an announcement sent to the house’s authors, Ross said she does not believe the paper’s bestseller lists accurately reflect what is selling across the country.

In a letter sent over the Labor Day weekend, Ross explained her position:

“Increasingly, it appears that the Times has gathered book sale data in a manner which prioritizes liberal-themed books over conservative books and authors. The net result has been a bestseller list that has increasingly become less relevant to the Regnery audience, and less reflective of which books are actually selling best in the country, regardless of one’s political persuasion.”

Ross said Regnery has had numerous books appear on paper’s lists, including titles that have reached #1. But, she claimed, Regnery should have had more books hit the list—and chart higher on it. The last straw for Ross was when, for the week ending Aug. 23, The Big Lie by Dinesh D’Souza was #7 on the Times‘ adult nonfiction list “despite being #1 in sales out of all 15 books featured on the list, according to NPD BookScan,” Ross said. (NPD BookScan tracks roughly 85% of print sales in the U.S.)

While Ross believes that the Times list does have a liberal bias, she is also concerned with its overall methodology. She said she has complained to the paper about its bestseller lists, voicing her concerns about books she believes should have been on the list but didn’t make the cut. She was told, she went on, what everyone who asks the paper about its bestseller lists is told: their methodology is a secret. Given this response, Ross said she believes the paper may cherry-pick sales from stores that are in liberal areas.

In a statement released to the Associated Press, Times spokesperson Jordan Cohen said: “Our goal is that the lists reflect authentic best sellers. The political views of authors have no bearing on our rankings, and the notion that we would manipulate the lists to exclude books for political reasons is simply ludicrous.”

Ross, though, believes otherwise. And what particularly frustrates her, she added, is the fact that consumers don’t understand how the Times‘ compiles its bestseller lists, and the fact that they are based on a cross-section of sales from undisclosed stores. With this in mind, Ross has changed the benchmark for bonuses in Regnery’s author and employee contracts, switching the benchmark from the Times’ bestseller lists to the Publishers Weekly bestseller lists, which are powered by BookScan.

In addition to changing its bonus contracts, Regnery will switch all of its marketing materials to feature the PW bestseller lists.

“There is a principle at stake here,” Ross said. “We want to work with a list that reports accurate data.”