Josh Patashnik is probably looking at the wrong metric when he examines self-described liberals and conservatives -- party affiliation is probably a more accurate determinant of which columnists you agree with.
But his post reminded me that one possible explanation for the apparent overrepresentation of conservative columnists is that liberals are disproportionately found in big cities, where a paper like the New York Times will serve a few million of them, while conservatives are disproportionately found in rural communities, giving them more papers even as those papers don't serve many readers each. It's sort of like those maps where Bush won 98% of the country's square feet while getting only 51% of the actual vote. So it would be natural for papers in conservative communities to run more conservatives, just as The New York Times runs mostly liberals. There's no way of running this study, but in some ways the correct metric would be how well-read columnists are rather than how many papers they appear in.
Update: Whoops, the report addresses this by totaling up newspaper circulation. They find that “in a given week, nationally syndicated progressive columnists are published in newspapers with a combined total circulation of 125 million. Conservative columnists, on the other hand, are published in newspapers with a combined total circulation of more than 152 million.”
