Academics and pundits have showered attention on Left Turn, a provocative work by an insouciant, politically provocative economist. Criticism and defense focus on Tim Groseclose's methods—rightly so, given that the bulk of the book is on the novelty of his framework and the force of his evidence, and given the need to overcome resistance to both technical claims (“the media lambda is a universal constant”) and startling findings (RAND is a considerably more liberal source than the ACLU). I want to ask a different question. If all the elements of Groseclose's design—“political quotient” and “political slant,” indirect measures of bias via references to think tanks and advocacy groups, the assumption that the median member of Congress and the median voter are in the same position (see Puglisi and Snyder Reference Puglisi and Snyder2011), the parallel between news citations and citations by members of Congress, the variable duration and time period of data collection for the media outlets studied—if they combine to support the conclusion that media bias alters voters' attitudes and behavior and moves them to be more liberal, what is at stake? We have well articulated grounds for thinking that a disjuncture between voters' interests and opinions and those of their own elected representatives and political representation overall is a problem for democracy. Groseclose's argument rests on a parallel claim: if we treat the press as a fourth branch of government, and “one of the actual three branches were as unrepresentative of the people as the press is … such a situation would produce a revolution” (p. 100–101). This provides a bit of a window onto Groseclose's understanding of the problem of media bias, but I will argue that his view of representativeness rests on an odd “original position.”
Before turning there I offer two observations about the author's temperament and the temper of this book. Left Turn reads as a paean to contemporary economists and, to a lesser extent, empirical political scientists (for example, “NOMINATE is the most impressive research achievement ever in political science,” p. 141). Groseclose cites awards and endowed chairs, ranks journals (informing readers that the editors of Quarterly Journal of Economics [QJE] who published the original article from which Left Turn derives are based in the Harvard Economics Department), repeatedly invokes the New York Times' “young economists to watch” list, identifies “rising stars” and elite programs. I stopped counting the testimonials to colleagues; to give a taste, Groseclose introduces a conversation with John Zaller by identifying him as “possibly the best and most esteemed political scientist on the planet” (p. 178). Is this a welcome change from the churlish tone we find in a lot of academic writing? Or, as Groseclose leads us through supportive conversations and e-mails from these luminaries, is he garnering reflected glory and added authority for his work? On reflection, I reject my cynical first thought, for what emerges from Groseclose's laudatory sketches of researchers who contributed to his thinking is a picture of the life of academic research—of false starts and settling on methods and measures, of challenges and improvement owing to the generosity of colleagues. It is also a brief for the view that meritocracy prevails in social science.
A second notable thing about the temper of Left Turn is Groseclose's confessions and demonstrations of his own conservative ideology and political behavior, emotional accounts of his experiences of marginalization, and first-hand observations of the redefinition of extremism in the homogeneous liberal university environment where he resides. These anecdotes provide motivation and personalize the book. Beyond that, Groseclose enacts the injunction to transparency he urges on errant, biased journalists. Does insistent political self-presentation really help readers judge the objectivity of his study, though? If we consider that the study is restricted to news reporting and not talk radio where conservatives hold sway, and that it does not cover the period of Fox News trumpeting, labeling, mobilizing, indeed orchestrating the Tea Party (see Skocpol and Williamson Reference Skocpol and Williamson2012), does Groseclose's embattled self-presentation have the unintended effect of provoking resistance to the significance of his findings? Does it increase the self-discipline readers require to focus on his 20 selected news outlets and to disregard the simultaneous effects of pundits, advocacy groups, talk radio, partisan blogs, paid political advertising, and smart-phone prompts that he does not put in the balance on the conservative media side?
I turn now from the tone of Left Turn to its argument, viewed from my standpoint in political theory. Groseclose has a normative thesis, though it is not explicitly defended: media news reports should mirror the “natural” political position of American voters. “The political views that we currently see in Americans are not their natural views” (p. vii), and these left-leaning deviations are the effect of biased reporting. There is a point on the liberal/conservative spectrum where the average American voter would be and therefore should be were it not for bias. Groseclose does more than uncover slant and its effects, then. Though he says little about the origin, time-frame, and persistence of liberal bias or the mechanism by which media bias alters political views, he does judge. The statement “an area founded by Confederate sympathizers and filled mostly with Mormons … is still not as conservative as Washington correspondents are liberal,” is accusatory (p. 125). But why would congruence between news reporting and voters' attitudes or ideology be desirable? The significance of this work depends not only on demonstrating bias and its effects but also on some even very thin account of democracy that explains why concerned citizens (especially liberal partisans) should worry about it.
For conservatives, of course, the answer is clear: media bias “aids Democratic candidates about 8 to 10 percentage points” (p. 245). Absent slant, “the natural views of the average voter” would be more conservative, positioned at a Political Quotient (PQ) of 31.5 rather than 50.4 (p. 239). She would share the politics of NASCAR fans. She would think and vote like Bill O'Reilly rather than Senator Olympia Snowe. Deflecting individual attitudes from their natural position alters voting and in the aggregate skews political outcomes and policy.
Notice that in the study design and his broader discussion Groseclose privileges “Independents.” He locates the media vis-à-vis this group, explaining, “Since I am most interested in how the average voter responds to news, I focus only on Independents” (p. 192). In the course of discussion he seems to refer interchangeably to Independents, the average voter, the median voter, centrists, and “moderates”—an assimilation of categories that goes unexplained. Presumably, they all support the point that, because elections are decided by Independents, he can set aside the effects of news consumption on partisans as politically less relevant (though this is not the case everywhere or consistently over time, and in most cases is open to dispute). In the same vein, he advises reporters who want influence to write to this audience. A better reason for focusing on Independents may be that Groseclose thinks they are more open-minded or that they process information differently so are more apt to be changed by a reduction in liberal reporting than partisans whose views are entrenched.
So, what would count as unbiased media? Consider three possibilities. Groseclose touches on two of them, but in the end proposes an alternative view of his own that I call his “original position.” One account has it that bias is ineliminable, and to mitigate its effects news conveyors should publicize their political attitudes and behavior. Groseclose offers his website for self-reporting; he will post how reporters vote and measure their “political quotient.” He insists that if a journalist feels that airing her political views is not part of the job, “she should consider that part of her career is based on deception.” Recognizing too that reporters may suffer false consciousness or sheer ideological obliviousness about where they fall on the political spectrum, Groseclose invites them to visit Dixie Gun and Fish in Washington County, Utah, to achieve self-awareness. As a way of mitigating bias, this position depends on journalists' awakening. More importantly it rests on the thought that transparency produces information that might offset bias. We know from the campaign finance context among others that without interpretation, publicizing information about donors' ideology and special interests does not have much impact. Scholarship in political psychology (which makes surprisingly little appearance in this book), particularly work on how we process new information, suggests other reasons that militate against unbiased evaluation of reporters' confessions, and therefore against purported beneficial effects.
A second way of thinking about unbiased media has it that every news article, television, and internet report should be “balanced.” Journalists should report as if they were Independents. They should offer countervailing facts and countervailing political interpretations of facts, cite Hoover Institute reports as well as Brookings, interview both (or all) “sides.” On this view, neutrality or unbiased reporting is neither a transcendent view from nowhere nor synthetic but a matter of setting out existing positions. This corrective too is limited in obvious ways. Requiring selection and insight, it depends on journalists' knowledge of the political terrain issue by issue—as well as adherence to this professional norm. It does not address bias in the selection and amount of coverage of news stories or the unacknowledged use of politically loaded terms (to take a recent example, what determines whether a political action is described as rising to the level of a “movement?”). Balance is what journalistic standards typically enjoin and what reporters claim to do, though Groseclose illustrates as well as measures failings; his detailed account of a Los Angeles Times report on admissions policy at UCLA is a compelling cautionary tale. A standard of balance along these lines would doubtless stimulate data collection and monitoring and fuel the multiplication of watch-dog groups. But what about the effect on voters?
“Balance” is a classic value in theories of democratic education. A considerable amount of time and money has been invested to test the effects of providing citizens with improved information in scrupulously designed and controlled settings such as citizen juries or deliberative polls. These experiments involve either randomly selected participants or Groseclose's Independents. They follow some variant of the objective set out in deliberative democratic theory: to encourage political decision-making based on reason-giving. The assumption is that on just about any issue citizens have too little information and erroneous information as well as biased information; that their understanding of political events, institutions, cause and effect, and likely consequences of policy are inaccurate; that they are underexposed to political views in opposition to their own and reasons for holding them. Information (often presented by moderators) must be selected for balance, completeness, and salience. It does not follow from effects in controlled settings—where the changes recorded are often in a progressive direction and the results of deliberation are not always moderating—that the outcomes would be similar for individuals outside of the group setting. Quite the contrary. Nor does it follow that if reporting were similarly selective and balanced citizens would access and weigh these sources over “The Daily Show.” Or if they did, that they would use the information and perspectives on offer in the manner hoped for. In any case, Groseclose mentions but does not set too much store on balance. His subject is not political ignorance generally or incomplete information or slight grasp of the logic of argument, or forms of rational irrationality. His subject is not the quality of information and interpretation overall. He is not pressing here for more enlightened partisans or Independents, but for more “natural” ones.
A third possibility has a distinguished pedigree in the history of democratic theory: the Millian dynamic that pits argument against argument, narrative frame against narrative frame, reasons against reasons, partisan against partisan. On this view, the clash of men and positions in a system of organized disagreement, and not internally balanced reports, is the mechanism for overcoming error and prejudice and achieving “truth and accuracy.” The logic of this dynamic aside, it has a sort of intuitive persuasiveness: our attention is gripped by conflict and close races, by “the rough process of a struggle between combatants fighting under hostile banners.” A contest waged not by honorable Independents either, Mill explains, but by “persons who actually believe them, who defend them in earnest, and do their utmost for them.” There is nothing wrong with a partisan press, on this view; the system of opposition is the corrective. Given the small pool of conservative journalists and what he sees as the absence of a vigorous conflict of opposing views, it is not surprising that this is not Groseclose's proposal for mitigating the effects of bias.
Instead, he presents his own account of unbiased media: reporting that mirrors the attitudes of the median (or average, or moderate, or Independent) voter. “Imagine a world where journalists report the news as the average voter wants them to report it … this means that she adopts a slant quotient equal to the political quotient of the average voter.” Groseclose does not mean “the whims and beliefs of the average voter,” either, but truth and accuracy in reflecting their stable views (p. 54), which ups the ante on mirroring where Americans actually are and the reporting they presumably want. There is a political reality as perceived by most Americans, he insists, and news reporting should reflect this reality back to readers and viewers. Reporters should provide “clarity.”
Again, what democratic value underlies congruence as a standard? The demand for mirroring might be grounded in sheer democratic indignation at the often covert exercise of power by media elites. On this view, liberal domination of reporting is domination (or manipulation) whether or not it is allied with corporate America and even if it is decentralized. In the election law literature we find indignation at one party, or in some case two-party, duopoly controlling outcomes through districting, and a demand for trust-busting; control of slant in reporting is its counterpart. The significance of congruence as a standard on this account is that it provides a rough baseline for demonstrating that ideological forces are unduly influencing citizens' views and political outcomes.
But Groseclose doesn't stop here, for mirroring actual (i.e., “distorted”) positions still effectively reinforces bias. Even so, he argues, that mirroring current, especially moderate, attitudes would push in the right direction, toward “natural views.” But why should we think that the period of his study is a source of the right baseline, the true, natural, conservative resting place? What if an earlier period was marked by conservative media bias, or by news reporting that reflected the position of the general public? Groseclose's “original position” is a calculation from within the period of his study and system of measurement, and this calculation is entirely contrived. It is outside the events of political history. It is isolated from other social forces that operate at the same time to shape political views. It is remote from the self-understanding and the encumbered political identity of real people. We need normative reasons, grounded in democratic theory, to accept mirroring as a standard for reporting and a return to our “natural” pose as a goal. In a rare, explicit statement of a democratic value underlying his thesis, Groseclose states that “the median voter in the United States should get his way with policy” p. (58); in fact, his concern goes beyond deviation from even that standard of political representation. It extends to distance from the “natural, non-distorted position of the average voter” (p. 6), i.e., the perspective of Groseclose's “ideal” voter.
In democratic theory, the media is valued for its part in the formation of opinion; it stands alongside other social forces and institutions that have an explicit mandate to shape political values and to educate “the American mind.” Citizens' political views are a selective reflection of what is presented to them, by way of processes political psychologists seek to analyze. Groseclose reverses this: the media should reflect where Americans are and should report the news as the average voter wants them to report it. He is not only a conservative partisan politically, he is a partisan for media deference, and an opponent of media activism (except insofar as we depend on the biased institution to correct itself). The core of this book lies here, in the shift from democratic education as improving and aspirational to the odd and unjustified imperative to return the “real” median voter to her hypothetical but authentic original position.
