August 7, 2003

Why Mr. Tomasky gets an F…

I want to spend a little more time today considering the recent study on partisanship in newspaper editorials by The Joan Shorenstein Center for Press, Politics, and Public Policy at Harvard University. Among the findings: “…conservative editorial pages are far less willing to criticize a Republican administration than liberal pages are willing to take issue with a Democratic administration.”

Yesterday, I called this result “intuitive” because “ideology is not experienced as an intellectual choice made by individuals. Instead, ideology is experienced as a self-evident moral system. The self-evident moral system of conservatives is generally more cohesive and prescriptive than that of liberals. So we should expect exactly these results.” For background regarding these assertions, please refer to this list of past blog entries.

You should also read Jay Manifold’s criticisms as a prelude to my own.

Today, I will demonstrate why this study is seriously flawed and what that means.

Let’s start with this section of the introduction:

One often hears it said that Democrats have The New York Times and The Washington Post, Republicans have The Wall Street Journal and The Washington Times, that the papers serve similar functions for each side, and that it all pretty much evens out in the wash. Does it? This is the question my research will attempt to answer. In this hyper-partisan age, are all editorial pages created equal-are the degrees of partisanship of the liberal newspapers and the conservative ones essentially the same? It is an important question because if one accepts the terms laid out above-that editorial pages signal the parameters of the permissible to decision-makers-then surely what each side considers permissible is a question with relevance to both the making of policy and the way day-to-day politics plays out in Washington. How strictly do the liberal pages follow the Democratic line? The conservative pages the Republican one? How capable is each side of criticizing its own? And how much credit is each capable of giving, however grudgingly, to the other side?

The writer, Michael Tomasky, seems to misunderstand how ideology works. There is no universal standard of civic discourse or the dialectically permissible. An editorial springs from an ideology. That ideology is experienced as having the force of reality, or self-evident truth, to those who hold it. As I stated above, the self-evident moral system of conservatives is generally more cohesive and prescriptive than that of liberals (Re: George Lakoff’s “Moral Politics”). It doesn’t take a study to assert that liberals will tend to allow a wider range of ideological flux in their discourse. That is neither good nor bad.

Tomasky demonstrates his own ideological bias–a liberal bias–for a wider range of ideological flux in editorials and asserts this as an ideal. So his research question is flawed by his own inability to see his inquiry as ideological and, therefore, partisan.

Continuing from the introduction:

Ours is an era of choleric argument about degrees of partisanship and bias in the media. Conservatives have spent years denouncing “the liberal media,” while liberals more recently have started to take up the opposite chant. Editorial pages, of course, constitute one journalistic venue in which bias is acceptable, indeed the rule. As such, the degrees and intensity of their partisanship can give us insights into how each side experiences and expresses its partisanship; that, in turn, can tell us much about the nature and parameters of political debate in contemporary Washington.

I find this interesting. And I agree with his final assertion. What this tells me is: This study should have been a proper rhetorical analysis, not an attempt to quantify partisanship by generalizing from limited textual data of dissimilar political situations (re: Manifold’s criticisms).

From the methods section:

I chose ten “roughly comparable” incidents from the two administrations and looked at how editorials in each of the four newspapers wrote about those incidents. “Roughly comparable” incidents are defined as issues or controversies on which there was a rough equivalence between what the Clinton administration was trying to do and what the Bush administration was trying to do.

You know you’re in deep academic trouble when your stated method is based on an obvious tautology.

More from the methods section:

Then, once all the editorials were assembled, they were given one of three designations: positive (P), mixed (M), or negative (N). A positive designation meant the editorial supported the administration

8 Responses

  1. markus 

    How strictly do the liberal pages follow the Democratic line? The conservative pages the Republican one? How capable is each side of criticizing its own? And how much credit is each capable of giving, however grudgingly, to the other side?
    The writer, Michael Tomasky, seems to misunderstand how ideology works. There is no universal standard of civic discourse or the dialectically permissible. An editorial springs from an ideology. That ideology is experienced as having the force of reality, or self-evident truth, to those who hold it. As I stated above, the self-evident moral system of conservatives is generally more cohesive and prescriptive than that of liberals (Re: George Lakoff’s “Moral Politics”). It doesn’t take a study to assert that liberals will tend to allow a wider range of ideological flux in their discourse. That is neither good nor bad.
    You get an F (in an otherwise fine post (except for the tautology charge which I consider invalid)) for insufficient explantion here. IF the ideology of the right is indeed “right or wrong, my party” such an ideology can objectively be considered “dumber” than the alternative, because it contains a disconnect from/disregard of the facts and thus reality. I know this is not really your position, but merely saying all ideologies are equal won’t do IMO.
    There is solid evidence that a “cohesive and prescriptive” climate of discourse leads to bad decision making (e.g. groupthink). Of course the other extreme is just as bad. However, it seems from the study that the NYT and WP are a bit too lax, while the WSJ and WT are defintively rigid in some instances.
    Care to elaborate?
    yours

  2. acline 

    I do not think all ideologies are equal. I believe what I said about what liberals would tend to do is, for the moment, accurate. It is neither good nor bad *until* we apply that assertion to socio-political situations and see what we come up with (always done ideologically, btw). That’s beyond the scope of this post.

    Ideologies are lived experiences. As such, they appear to be truth to those who experience them. We may eventually judge an ideology flawed or worse, but that doesn’t change the experience for the individual holder of the ideology (unless they are persuaded otherwise by further experience or a damned good argument).

    I’m fascinated why you would think the tautology “charge” is invalid. Here’s the quote: “‘Roughly comparable’ incidents are defined as issues or controversies on which there was a rough equivalence between what the Clinton administration was trying to do and what the Bush administration was trying to do.” Hmmmm…this is pretty clear to me. “Roughly comparable” equals “rough equivalence.” The elaboration, IMO, at the end of the sentence does little to clear this up. One big reason: What presidents are trying to do is always politically and ideologically driven. To make such a comparison requires that we find an equivalence that, I would argue, doesn’t exist between two men as different as Bush and Clinton.

    I’ll take my ‘F’ begrudgingly 🙂 I should know better than to try to cover so much territory in a blog post. Oh well.

    Thanks for you comment!

  3. markus 

    the tautology first: i agree the definition is rather loose, personally I would have preferred a rating by experts not involved in the study. However, the Tomasky states that roughly comparable incidents to him are incidents with whose goals have a similar impact on the country (my paraphrase). He did not take the number of respective editorials, the amount of taxpayer money involved or the result of the initiative (easy versus hard victories/failures) as his standard of rough equivalence. That opens him to criticism, such as that Clinton’s first year initiative involved less taxpayer money and thus is not comparable to Bush’s enormous tax cut.
    From where I’m standing his specification of “roughly equivalent” is not satisfactory, but since it excludes other standards of comparison and “rough equivalence” it is not tautological.

    Concerning ideologies: Thanks for clarifying that your statement on the neutrality of ideologies primarily refers to out of context judgement (“It is neither good nor bad *until* we apply that assertion to socio-political situations.”) I agree with your statements concerning the self evident nature of ideology. Still, I’d argue, that Tomasky is investigating applied ideology here, and so judgement is possible. In so far, as I fail to see a difference between a more cohesive and prescriptive ideology and an increase in partisanship I believe Tomasky’s judgement is acceptable, provided we can agree that partisanship in editorials is negative.*
    It seems you disagree on the grounds that the aforementioned judgement is influenced by ideology. My point (you seemed to agree in your reply) was and is, that the mere fact of ideological influence on decision making and value judgement does not invalidate the judgement in itself. Facts and arguments count. Tomasky provided one such argument, namely the historical concept of journalistic independence (which also justifies certain special rights for journalists). You provide a counter argument, based on your perception of the purpose of editorials, namely to persuade. Fair enough, I’ll just counter with my own perception that the purpose of editorials is to provide ideas and arguments, which the reader is free to accept or reject. Maybe your view is grounded in some expertise on your part and thus more correct, I can’t tell from the post. But as it stands, it seems mere opinion, whereas my (and Tomsky’s view) is mere opinion based on historical concepts of journalism.
    Therefore, I feel it is on you to demonstrate the equal validity of your view before you can conclude that both views are equally in-/correct and thus cancel each other out (which is the basis for dismissing Tomasky’s judgement AFAICT).

    I apologise for the poor wording and hope I have made myself understandable.
    yours
    markus

    * even if we can’t agree, the finding that the WSJ and WT are more partisan (regardless of whether that’s good or bad) is relevant to the media bias debate.

  4. acline 

    You note an important distinction that I often forget to explain:

    Re: “You provide a counter argument, based on your perception of the purpose of editorials, namely to persuade. Fair enough, I’ll just counter with my own perception that the purpose of editorials is to provide ideas and arguments, which the reader is free to accept or reject. Maybe your view is grounded in some expertise on your part and thus more correct, I can’t tell from the post. But as it stands, it seems mere opinion, whereas my (and Tomsky’s view) is mere opinion based on historical concepts of journalism.”

    First, I’d rather not deal in issues correctness. I’d rather ask: What works? That said, I am a rhetoric scholar. I operate from a theory that suggests all human communication is, at its base, rhetorical, i.e. an attempt to persuade. There are certainly theorists who disagree.

    One who operates from such a theory would naturally understand newspaper editorials as attempts to persuade. The following is NOT directed at you: I think journalists often kid themselves about their persuasive intents, e.g. “This is just something to think about. You make up your own mind.”. And, not being particularly well trained in language issues beyond questions of style, most journalists haven’t a clue about the disciplines of rhetoric or linguistics or any of the language theories that attempt to explain what it is we do when we communicate with one another. (I was a journalist for about 15 years, so I’m speaking from pre-academic experience.)

    This is part of my problem with Tomasky (and, really, with Harvard’s PPP program). I think his essay lacks proper academic grounding in theory. Where’s he coming from? I suspect he doesn’t know.

    Is partisanship bad in newspaper editorials. I suspect that it is. But I think I understand why its there in differing degrees among the papers Tomasky studied. Rather than give me an equally partisan conclusion, I would rather have seen him posit a theory that *explains* this phenomenon. Then, I think, we can properly begin to decide if its good or bad.

    Here’s just one way he could have handled it: Lakoff and Johnson’s work in metaphor could be useful here–especially Lakoff’s work in the moral ramifications of political metaphor. An interesting rhetorical analysis of those editorials could have used their theory of metaphor to demonstrate why the differences in partisanship exist. Then, piggybacking onto Lakoff’s conclusion from “Moral Politics,” he could have used that theory to properly assert an evaluation.

    I appreciate your intelligent comments so far!

  5. markus 

    That said, I am a rhetoric scholar. I operate from a theory that suggests all human communication is, at its base, rhetorical, i.e. an attempt to persuade. There are certainly theorists who disagree.
    I think I understand, but remain committed to the latter group. AFAIC in my discipline (psychology) your approach has not been very successful (which certainly says more about psychology than about rhetorics). I’m working a few conceptual levels below you though (visual word recognition), so my assessment might be wrong.
    concerning journalists we are in perfect agreement, though my pet peeve is their poor understanding of statistics. 😉
    concerning your suggestions for improvement, I wholeheartedly agree, but -at the risk of being stupid or stubborn- I’d argue that this is not the study Tomasky set out to do. Re-reading the post and our contributions just now (which certainly were attempts to persuade) I think we’ve been discussing two different studies. From your quote of the paper:

    …papers serve similar functions for each side [your study], and that it all pretty much evens out in the wash. Does it? This is the question my research will attempt to answer. In this hyper-partisan age, are all editorial pages created equal-are the degrees of partisanship of the liberal newspapers and the conservative ones essentially the same [my study]? … How strictly do the liberal pages follow the Democratic line? The conservative pages the Republican one?[your study] How capable is each side of criticizing its own? And how much credit is each capable of giving, however grudgingly, to the other side?[mostly my study]

    To me, the study was about quantifying and demonstrating the intuitive notion that partisanship is more intense on the right. This is a perfectly acceptable point of enquiry for me, as much of psychology was concerned with checking intuitive notions. Nowadays of course a theory is mandatory, but I’m not convinced this requirement is always beneficial to the progress of my discipline.(whatever your position, I’ll accept it, please don’t start an argument on this last sentence.)
    Anyway, when I try to take your point of view, I completely agree that the study lacks theoretical grounding. (Again, what is the real conclusion to you is the free-speculation-after-a-hard-day’s-work to me. Psychologists do it all the time.) The study merely offers a layman’s theory, and I believe we were arguing over different things again. I argued for the basic correctness of the layman theory, you argued that the layman theory is insufficient. FWIW, I believe you’re right.

    Provided my interpretation of our exchange is correct, I am beginning to wonder whether you pulled some Socratic dialogue thingy on me, small steps and all that. Anyway, it’s been enlightening*, so I’ll gladly return the compliment.

    yours
    markus
    * leaving out the “so far” bit, as I don’t expect us to descend into name-calling at this point.

  6. There are two different studies here because we are two different people. 🙂

    I think you’re correct that “the study was about quantifying and demonstrating the intuitive notion that partisanship is more intense on the right.” In that regard, I think it could have been handled better if Tomasky had been a bit more self-conscious about his theoretical assumptions. He states and/or infers his assumptions as self-evident facts, which leads me to believe he hasn’t a clue about them.

    As for the Socratic thing 🙂 well…I don’t do that on purpose, the question-answer thingy that is, because I consider that method intellectual mugging. But I do tend to baby-step it a bit if for no other reason that I think discussions, especially in an online textual environment, can get quickly off track to the detriment of any real understanding among all of the participants.

    I look forward to more of your participation here!

  7. Rhetorica does the heavy lifting

    I pointed out the new Harvard study on media bias yesterday – remember, the one where the new editor of…

  8. Sean Landis 

    Dr. Cline,

    You might be interested in Bob Garfield’s interview with Michael Tomasky on On the Media (transcript here: http://www.wnyc.org/onthemedia/transcripts/transcripts_080803_slant.html)

    from the interview: I mean when a newspaper takes two roughly similar situations and writes such completely different things as the Wall Street Journal did in those two episodes, well I think that is dishonest. There’s only one explanation for it! They’re on one side and they’re against the other side. However in other ways, I admire [the conservative papers’] model a little bit more than I admire the New York Times and, and Washington Post’s model. Be partisan. Be aggressive. They’re certainly winning with it. [LAUGHTER] At least for the time being.

    Also, if you want to restart the debate on Tomasky’s report, Ken Waights of lyinginponds.com has an interesting take, including rebuttal to your analysis. (It’s here: http://www.lyinginponds.com/archive.200308.html#20030813)

    from Waights’ analysis: An editorial board might believe that a White House occupied by their own party was right 80% or 90% of the time on policy, but if they were not partisan they would certainly take note of human foibles which are universally condemned but common in all political operations — hypocrisy, hubris, fumbled nominations, a desire for secrecy (e.g. the Cheney energy task force) and so on. This is the key point which I think Andrew Cline misses in his critique of Mr. Tomasky’s report: “It doesn’t take a study to assert that liberals will tend to allow a wider range of ideological flux in their discourse. That is neither good nor bad.” But the major difference between the NYT and WSJ was not in the ideological range of their opinions, but in their approach to these less-ideological issues.