The Theoretical Case for Liberal Bias in the News
Evidence for empirical claims comes in two sorts: data about a particular case and data for a general pattern. For example, suppose I want to know whether the bird in my front yard is Baltimore Oriole. One thing I might do is go outside, take some pictures, and compare them with my guide book. The pictures provide me with evidence in the form of data about a particular case that I can use to confirm my hunch.
Another thing I might do is just to look up the normal range of the Baltimore Oriole. When I do so, I'd find out that they are rarely found on the Western Slope of Colorado. In a case like that, I wouldn't have any data on the particular bird that I saw in my yard, but I would have evidence in the form of a well-supported theory about the liklihood of an Oriole making it's way to my house. In this case, I have pretty good reason to doubt my hunch even in the absence of further information on any particular sample.
Here's how the distinction relates to news bias. There is a TON of research of the first kind confirming a left-leaning bias in many, many news sources (for example, the Ad Fontes chart). All of this research relies on data about particular cases. For example, social scientists collect samples of media, score them for liberal framing, topics, etc., and then add the scores up. This data collection paints a fairly consistent picture: news sources vary widely on political bias but there is a significant chunk of media sources that are regularly and significantly biased to the left. Details about the particular samples of reporting from places like CNN, The New York Times, Washington Post, MSNBC, etc. confirm that they have a clear bias to the left.
Despite the existence of this data, many of my liberal friends and colleagues deny that mainstream news sources are left-leaning. They often quibble with the methodologies or results of studies showing a left-leaning bias. In other words, many of them don't trust the data about these particular cases. As a result, they don't accept that many of our trusted news sources are biased towards the left.
This brings us back to the first sentence of this post: evidence comes in two sorts: data about particular cases and data about general patterns (theories). So let me offer a theoretical reason for thinking that many (most?) of our news sources are biased to the left. This reason holds even in the absence of data about particular sources to back it up. You don't have to gather any samples of news media or measure the bias of a news story in any way to accept the theoretical argument I offer here.
The argument has two premises:
Most journalists are politically liberal.
Your worldview shapes your perspective.
If you accept (1) and (2), then, even if you distrust all of the studies about particular sources, you have an independent, theoretical reason to think that the news will be biased to the left. Are both premises true?
As to the first, there is overwhelming and conclusive evidence that journalists--and especially journalists who cover political issues--are to the left of the American public (and in many cases, far to the left). For example, when we look at mandated disclosures about campaign contributions, journalists and staffers at news media companies regularly give about ten times as much money to liberal candidates as conservative ones (see here, for example). A 1995 survey of Washington-based reporters found that 90% had voted for Bill Clinton and only 7% for George H.W. Bush.
More recent social science data confirms this trend. Even in two studies that find no gate-keeping effect (that is, no correlation between a journalist's ideology and the stories she chooses to cover), both find that journalists are very liberal. A WashPo study found that 80% of journalists who identified with a political party self-identified as liberal/Democrats. A Science Advances paper shows that journalists are overwhelmingly more liberal than the American public; in fact, 1 in every 6 reporters is even more liberal than Representative Ocasio-Cortez, one of the most liberal members of Congress by any reasonable measure! Given all this, premise 1 seems like a gimme.
So how about premise 2?
One of the great things about a college education is that it shows you how much of your interpretation of the world is just that: interpretation. You take in evidence, experience, data, and much more. You then shape and push and pull that information into an interpretation of what's really out there. Your interpretation differs from mine. Your unique vantage point gives you advantages that I lack and vice-versa. It also hampers your ability to find certain truths that I have no difficulty discovering and vice-versa. We see the world from different views. The world isn't just given to us. We construct a map of the world, and that map requires plenty of creative thinking on our parts.
My worldview affects what I see (and what I don't) and what I think (and what I don't). Two people will radically different perspectives on the world are likely to disagree about what's true, what's important, what's right, and what's wrong.
In fact, this is one of the two really important reasons that higher education should care about demographic diversity among faculty. If your teachers are all white (or male or heterosexual or...), it's incredibly likely that important truths are being overlooked and important points downplayed. Who we are RADICALLY affects how we see the world. If we want students to get a more accurate or more complete view of the world, we better make sure that our teachers occupy a wide variety of perspectives. That's a strong reason for thinking that premise 2 is true.
So, the theoretical argument stands. Reporters and journalists are overwhelmingly liberal, and one's political perspective has a significant impact on how one sees the world. Given both of those things, it's probable that the reporting and stories produces by these individuals have a liberal bias (in topic, framing, language, etc.). That's a powerful, yet theoretical, reason for thinking that many news sources will have a leftward tilt.
Here's an objection you might have considered while following the lines of this argument. Isn't it possible to have a liberal (or conservative or religious or whatever) perspective and yet not allow that perspective to bias your actions? After all, we expect people like judges, teachers, etc. to practice their trades in unbiased ways even while holding a perspective that biases their outlooks. Why can't we just expect journalists to do the same? They might have their own perspectives, but they can check those at the door when they write the news.
This objection is right as far as it goes, but that's not saying much. Yes, we expect a wide range of people to act in unbiased ways even while acknowledging that each of them really has a perspective and bias. Teachers are perfect examples. I teach courses about ethics, God, politics, and a host of other controversial topics. I strive to be objective at every level. And yet I know that I inevitably fall short of that bar.
So, given that we know our limits, we implement a plan B. While continuing to strive for objectivity, we employ a wide range of people with a wide range of perspectives to the positions that require objectivity. That's a kind of meta-level panacea for a first-order problem of informational bias. If you know the bias creeps in, do what you can to cancel it out at the level of information production.
And, for my friends on the left, notice that nothing short of that level of perspective-diversity would satisfy you in domains other than the news. You wouldn't be happy with an all white faculty teaching courses on race with the encouragement to think about those issues in an objective way. You wouldn't be happy with a judicial system consisting of an all-male judiciary given the encouragement to think about gender issues in an objective way. By parity of reason, nothing short of political and viewpoint diversity should be accepted on the newsroom floor. Until that happens, consumers of the news will have a theoretical reason to think that the perspective provided by journalists and reporters will be skewed to the left.