Last week we solicited your questions for Tim Groseclose, a political science professor at UCLA and author of the new book, Left Turn: How Liberal Media Bias Distorts the American Mind. The response was fast and furious. A total of 149 questions (and counting) have been posted in the comments section. We selected 14 of them for Groseclose to answer, and he obliged us quite promptly. As always, thanks to all for participating.
Q. Why does liberal media bias exist in the first place? What would you suggest as a way that a) journalists could be more aware of their own bias and limit it in their reporting; or b) the profession of journalism could attract a more unbiased (or merely more representative) cohort? – Jack
A. The main reason why bias exists, I believe, is simply that newsrooms are filled overwhelmingly with liberals. Here’s the most important fact to know, if you want to understand media bias: If you poll Washington correspondents and ask “Who’d you vote for last election?”, about 93% will say the Democrat.
Why are newsrooms so liberal? I don’t know, except that I suspect that it’s mainly self-selection. I believe that there is something in the DNA of liberals that makes them want to pursue careers like journalism, academia, and the arts.
A manager or owner of a media outlet could try to counteract this by trying to hire more conservatives, but he will have a hard time trying to find conservatives who want to be journalists. He’ll either have to pay conservative journalists more or be willing to hire conservative journalists who are not as good at reporting as liberal journalists. It’s a hard problem for a news-outlet manager to solve. I basically believe we’re in an equilibrium – that liberal bias is basically here to stay.
How can journalists be aware of their own biases? One way is to read Chapter 11 of my book, “The Anti-Newsroom, Washington, County, Utah.” In the chapter I search for a place that votes the opposite of a newsroom – 93-7 for the Republican. It’s basically impossible to find such a county, but one that comes close is Washington County. I interview lots of people in the county to give the reader a sense of what political views in the anti-newsroom are like. If journalists think about how conservative, and maybe even strange, views are in the anti-newsroom, they may begin to realize how liberal, and maybe even strange, views are in actual newsrooms.
And if a journalist is really serious about understanding his or her own biases, he or she could visit Washington County, Utah. One of its residents, Tom Seegmiller, has agreed to host such journalists. Seegmiller is the owner of Dixie Gun and Fish and the Locker Room, an athletic supply store. If such journalists are interested, they should contact Seegmiller at one of his two businesses. Seegmiller is even willing to take such journalists to church with him. And if they desire, Seegmiller is willing to take such journalists hunting with him.
Q. How do you account for the filter bubble effect — that liberals and conservatives alike listen to media that doesn’t challenge their views? – Rachel
A. All my results about where people get their news involved surveys about where independents get their news. I ignored the results involving Republicans and Democrats.
Although Republicans and Democrats probably do get their news from vastly different sources, in one sense it hardly matters. To win a nationwide election, you need to win over the independents. Republicans and Democrats basically cancel each other out. To change policy, the key is to persuade independents/moderates.
Q. My question is this: is self-selection at work in media companies (both the liberal and conservative medias)? And, what other occupations have strong political self-selection? Are bankers more likely to be conservative? Are artists and actors more likely to be liberal? – Caleb b
A. Yes, I think self-selection is the key. But it feeds on itself. That is, once the newsroom becomes overwhelmingly liberal, it becomes less pleasant for conservatives. Consequently, conservatives become even more reluctant to become journalists.
As a conservative professor, I can speak from experience, that when people from one political group begin to dominate an organization, they can sometimes become a little sanctimonious and tedious. See, for example, page 4 of my book, where I describe an email that my co-author received from one of his fellow University of Missouri professors. (I think Amazon allows you to read the page for free.)
I think probably the most conservative profession is military officer. So I hear, military officers vote about the opposite way that journalists vote.
Q. What role does religion play in these biases? Michele Bachmann and Jim DeMint are both Christian ideologues, and while there are certainly “anti-religion” ideologues on the left, neither Barney Frank nor Nancy Pelosi would qualify. – Lawrence
A. I think you are probably right; there are no strong anti-religion ideologues in Congress. But I think the vast majority of Americans are fairly pro-religion. If America were divided 50-50 on religion vs. anti-religion, I believe you’d see more anti-religious ideologues in Congress.
But just because there are two sides to an issue, that does not mean that a reporter should give each side equal treatment. That is, “unbiased” does not always mean giving equal treatment to two sides of an issue.
For instance, lots of people (and I am one of them) believe that the evidence suggesting (i), that the earth is warming, is greater than (ii), that the earth is not warming. Thus, to be unbiased, I believe that a reporter should give more favorable treatment to (i) than (ii).
A hero of the left, Edward R. Murrow, may have made this point best. Interestingly, he used a religious example to make the point: To insist upon such an artificially equal treatment of two sides of an issue “is like balancing the views of Jesus Christ with Judas Iscariot.”
Q. Agreed; I think the strongest counterpoint to Mr. Groseclose’s premise is, what would the presidential split have been if all the influential media outlets weren’t owned by conservatives (re: Disney, Murdoch/Newscorp, GE, etc.) – cackalacka
A. I’m not sure I agree with the premise. If GE shareholders and executives are so conservative and have such power over their journalists, wouldn’t that cause Rachel Maddow and Ed Schultz to have a conservative bias? For much of the time that John Stossel was at ABC, the chairman of the Disney Board was George Mitchell – the former senator who’s PQ is about 80. If corporate executives are so powerful, wouldn’t we have seen a liberal bias from Stossel?
The New York Times is a corporation with two classes of shareholders. The class that has control over running the company contains a relatively small number of shareholders. The same is true with the Washington Post. I’m sure that with each company the shareholders are very liberal.
Meanwhile, the Washington Times is not a corporation (it is owned by the Unification Church), yet its slant is fairly conservative. So I’m not sure that it’s true that corporation-owned media companies tend to be more conservative than non-corporation-owned media companies.
To answer your question about the presidential election, suppose that for some reason all the media moved left – say all media began adopting a Slant Quotient of 74, like the New York Times. This would mean that the overall Slant Quotient of the media would move from 58 to 74, a change of 16 points. This would give Democrats an extra advantage of about 8 percentage points. Assuming everything else constant (e.g. Obama and McCain are still the candidates and they adopt the same policy positions as they did in the actual election), then, according to my results, Obama would have won by approximately 61-38, instead of the actual result, 53-46.
Q. And as a follow-up… if there exist institutions that provide a conservative bias, how do their ratings compare to one with a liberal bias? – Matthias
A. Well, I suppose that one of the most conservative groups in America is officers in the military. I’m not sure what it would mean to calculate a slant quotient for them.
Q. Given that the politics of the USA are significantly more conservative than most other developed nations, how applicable are your findings to an analysis of the mass media in other countries? – Brennan Young
A. Yeah, my book is completely silent on that question. I agree that other nations are generally more liberal than us. If I’m right, that there’s something in the DNA of liberals that makes them go into journalism, then I’d at least speculate that in other countries journalists similarly adopt a Slant Quotient to the left of the country’s average Political Quotient. But that’s just speculation, not evidence.
Q. Aren’t there other – perhaps more important – ideological axes than liberal/conservative? (e.g. statist vs. grassroots) – Brennan Young
A. Yes, there are definitely other axes. E.g. you could imagine a libertarian/anti-libertarian axis. But I’m not sure they are more important. Please see, for instance, my discussion on pages 40-44 of political scientist Keith Poole and the Nominate scores he created. Nominate estimates a numerical score for politicians on the “dimension of maximal conflict” within Congress. According to Nominate, politicians like Nancy Pelosi and Barney Frank are at one end, and politicians like Michele Bachmann and Jim DeMint are at the other. That is, the “dimension of maximal conflict” puts far-right conservatives at one end and far-left liberals at the other. It does not put libertarians like Ron Paul at one end and anti-libertarians at the other. Accordingly, Nominate suggests that the liberal/conservative dimension is indeed the most important dimension, at least in Congress. (This is not to say that in future years, things might change.)
Q. How does PQ vary by age and education? If more education is correlated with higher PQ, does that explain media bias since most journalists are required to have a college education? – Sam
A. It turns out that the people with the least education (non high school graduates) and those with the most education have the most liberal views. The voters with the most conservative views are those with intermediate amounts of education – those with only “some college” and those who completed a bachelor’s degree but did not attend grad school.
Q. How do you reconcile your conclusions with the fact that Americans appear to choose media you label as liberally biased when they have more conservative options? Why doesn’t the media reflect the supposedly conservative viewpoint of its consumers? Does the media really drive consumer thought in an open media market, or is the opposite true? – Ricky C
A. I think a lot of it is simply that it’s hard for a media organization to hire conservative reporters. I know that if academia suddenly decided “we need a balance of conservative and liberal professors,” then the next question that deans and department chairs would ask is “Okay, where do we find the conservative professors to hire?” I suspect something similar occurs with the media. Conservatives just don’t tend to want to enter into journalism, at least not at the same rate as liberals.
As a consequence, news outlets can hire liberal reporters at a lower wage rate than they would have to pay if they insisted on hiring conservatives. They can probably also get higher quality reporters if they’re willing to hire more liberals, simply because the pool of liberal reporters is larger than the pool of conservative reporters.
As I understand, a similar issue arises in baseball. Teams want a balance of right- and left- handed pitchers. But the pool of right-handed pitchers is much higher than that of left-handed pitchers (since there are much more right-handed people in the population than left-handed people). As a result, on lots of objective measures – e.g. throwing speed – right-hand pitchers tend to be better than left-hand pitchers. I think something similar might be occurring with liberal and conservative reporters.
Q. American public opinion is fickle on important issues. It is hard for me to consider the average American voter the “center” when that center appears to be a sporadically moving target. When talking about a “center”, you expect something a little more stable even as it shifts. Journalists, more ingrained with the issues and needing to maintain integrity over time, would have more stable opinions. Do you look at shifts over time? Do you have a PQ moving average? How does this compare to journalistic PQ? – Ricky C
A. Well, events cause us to change our views. I know my views have evolved over the years. (E.g. I used to think that abortion was okay when the fetus is three months old. But having a kid and seeing a sonogram changed that view.)
Nevertheless, I’d argue that, at least over the last half century or so, the American center has been pretty stable. E.g. if you check page 50 of my book, you can see a graph of how the center (i.e. average PQ of American voters) has evolved. Between 1960 and 2009, it’s remained within the 47-58 range, and usually it’s been very near 50.
Although I have a few surveys over how journalists vote in elections, I don’t have much data about their PQs. Part of the problem is that journalists are so reluctant to reveal their political views. If it were up to me, they’d be more transparent about such things. (See, e.g., the epilogue of my book.)
Q. Of the actual voting public what’s the percentage of viewers that get their news exclusively from the liberal media? What’s the percentage that gets it from both liberal and conservative view points? What’s the percentage getting their news exclusively from conservative viewpoints? What prevents one set of viewers from changing their news consumption habits? Is selection of media source an indication of enlightenment? – Deron
A. My answer to the first four questions is “I don’t know.” As for the fifth question, yes, I believe that anyone who has a PQ under 20 (like me) yet chooses to subscribe to the New York Times (as do I) is enlightened. I’d say the same thing about anyone who has a PQ above 80 yet chooses to frequently read the Washington Times or frequently watch the O’Reilly Factor or frequently listen to Rush Limbaugh.
Q. Do you have PQ scores for economists? I’d like to cross-reference this with the economists’ track records over the last 10 years so I can decide whether your idea of PQ is poppycock. – Ben
A. No, but on page 112 of my book I review the work of Dan Klein (at George Mason University) and Christopher Cardiff (at San Jose State University). They have tracked the voting behavior of economists and other professors. They find, for instance, that in a typical presidential election economics professors vote about 2.8:1 for the Democrat. (In sociology the ratio is 44:1; in political science 6.5:1; in electrical engineering 2.5:1, and in finance 0.5:1.)
As for economists having poor track records, I might agree with you, at least when it comes to macroeconomics. I believe that in 200 years people will look upon the current state of macroeconomics the way we look upon blood-letting doctors of two or three centuries ago (i.e., that they had no idea what they were doing).
Q. During the Depression, the overwhelming majority of newspapers opposed Franklin Roosevelt. Yet he won 3 re-elections easily. Does that mean that without the opposition of the newspapers, he would have won even more easily? – Paul
A. My results suggest yes – the media really do influence the way people think and vote.