There's a lot of talk that the media is liberally biased. I think this is true to some degree, but until lately I knew of no way to prove or disprove the allegation. Then, I discovered a simple, concrete, non-subjective way to at least gather a small amount of hard data.

Here's what you do:

  • Open your local newspaper to the editorial section (in particular, the place where editorial cartoons are displayed).
  • Next, eliminate any editorial cartoons that have nothing to do with politics. If it's about some local road project, a jab at a sports figure or some other non-political topic, ignore it.
  • Examine the editorial cartoons that do have obvious political messages and decide whether it's a liberal or conservative ox being gored. This is usually quite obvious. I've done this for a while and I've seen only a couple of syndicated political cartoons that poked fun at both sides.
  • Keep a running tally for ten days or a month or so.

I don't know about papers around the country (but would like to hear), but our paper here in Austin, Tx runs about 3 to 1 mocking conservative causes over liberal ones. Thus, by this small measure, it is liberally biased.

Why this methodology? Several reasons: it's easy (takes about two minutes a day), it's not very subjective (cartoons are much easier to assign to the D or R column than articles, which are sometimes balanced and fair), and it's something completely under the control of the local paper (they have access to dozens of cartoons each day and can publish whatever ones they like).

What's your experience with your paper?


ikeleib says: The editorial page reflects the opinions of the editors and is not necesarily meant to have evenhanded coverage.

Exactly! Where else but the editorial page would you find the actual bias of the staff exposed for all to see?

The idea that they're just giving the people what they want doesn't hold up under much scrutiny: they endorsed Bush for prez, and most Austinites voted for him (local boy and all), but now they now lambast him editorially much more than liberals or than they did Clinton, even in the depths of the Lewinsky scandal.

You have to wonder if they're not just a little touched by the heat.

Before we can plan such an experiment in any rigorously scientific sense, we must understand what "liberal" is. There lies the difficulty. There are a huge variety of issues that a newspaper may be biased on; how do we resolve the various positions on these issues into the category "liberal" or "not liberal" (in the first writeup, assumed to be "conservative")? Is a labor leader, in favor of raising the minimum wage but decreasing environmental controls, a "liberal"? Would some hypothetical scientist, in favor of increasing environmental controls but decreasing the minimum wage, also be a "liberal"? The political "spectrum", unfortunately, does not lend itself easily to a one-dimensional analysis like the one desired.

Let us say that we were capable of categorizing an arbitrary set of positions on different issues into "liberal" and "conservative". Could we then categorize a politician as "liberal" or "conservative", only on the basis of those views that the politician has made public to us? Assuming we could, does the fact that the newspaper targets the politician as cartoon fodder make the newspaper the "opposite"? A cartoon could lampoon a person who is ostensibly conservative for not being conservative enough, for example. How much subjective judgement do we allow ourselves as experimenters in deciding what the intent of the newspaper was? Do you "just know" intuitively what the cartoonist's intent was? How is this different from "just knowing" what the bias of the newspaper is?

Lastly, is the use of the political cartoons the best measurement? Presumably, one is interested in the effective bias- that bias that has some potential to sway the reader. Isn't the bias of a newspaper much more subtly persuasive in the use of affective words in the supposedly objective reporting of the main pages? However much our news sources give us the denotative truth, how much are we subconsciously led by the connotative aspects of their word choices? When one reads opinion pages, one is on guard, ready for the subjective. When one reads the "news", is one still on guard? Or does one take the newspaper's word for it?

While it is undeniably simple and less subjective than some methods might be, there is a crucial factor that shokwave's experiment overlooks: The relative ease of poking fun at the two sides. For instance, if George W. Bush gets parodied more than Bill Clinton did, is that necessarily because the newspaper or cartoonist is liberal? Or could it just as easily be simply because he is exceptionally ripe for satire? You may find that the difference in satirisability between 'liberal' and 'conservative' targets is substantial and pervasive, creating a systematic error in the data from this experiment which should not be ignored...

Some better experiments to divine the true bias in the US media would be:

Experiment One: Discover Your Corporate Masters!
  1. Cut out all of the advertisements. Put these in Pile One.
  2. Cut out anything that looks like an unedited press release from a corporation, and the entire Business Section. Place these in Pile One.
  3. Cut out anything that you think is "unbiased reporting" like the Weather. Place these aside for Experiment Two. If these are actually unbiased, they will have no effect on you.
  4. Try to find anything liberal in the remaining shreds of paper. Place these in Pile Two.
  5. Gaze awestruck at Pile One and wonder quietly what happened to the free market of ideas that capitalism provides. Contemplate whether liberal bias is the real issue.
Experiment Two: Episodic or Thematic?
  1. Now that the paper is cut into tasty bite-size pieces, divide news into that which is episode oriented - that which focuses on concrete issues such as an individual welfare mother, a crime spree, or a terrorist bombing. Label this Pile One.
  2. Next find the news that provides information that contextualizes an issue or problem, and focuses on general developments, trends, or preconditions that contribute to problems such as surges in welfare dependence, crime or increased terrorist attacks. Label this Pile Two.
  3. Gaze awestruck at Pile One and wonder quietly why Americans can't see the ongoing, systemic problems in society. Contemplate whether liberal bias is the real issue.

Shokwave's experiment proves only one thing: that cartoonists prefer to pick on the country's leaders than the also-rans of the opposition. A better exercise would be to take an article and note down all the facts that have been related therein (preferably checking them against another source). Now show the sheet of facts to someone who has not previously heard of the story. Ask them for their personal feelings on the subject. Then show them the article, and see if this has changed their perception of the issues.

Whether this experiment is scientific or appropriate to the question is irrelevant. The point is to show you what you should be doing every time you read, see, or hear an item of news: discerning the facts from the bullshit.

Log in or register to write something here or to contact authors.