There is a huge difference between public perceptions of the power of media in elections and academic evidence of its influence. This gap stems from the fact that the public uses different forms of evidence than academics use to infer media power. This essay outlines the reasons for this great divide, then highlights the seriousness of its consequences for the allocation of political resources. Public beliefs in omnipotent media contribute to wasted time and money; ultimately, they undermine the legitimacy of election outcomes.
As a scholar of media and politics, I am frequently asked to participate in media commentary during election years. Although I wholeheartedly believe in the importance of academic outreach to the larger world, I suspect that this is my least favorite part of my job. It is in this context that I am most often told – strongly, unequivocally, and unanimously – that I am wrong. The multitude of observations involving media and politics about which I am wrong is both wide and deep. They converge around my relative naiveté in understanding the sheer power of the monster. When I take part in a radio call-in program or appear on an election-night television broadcast, then I, too, become part of the monster, wielding its incredible power while simultaneously denying its very existence.
While both the public and academics agree that media have influence in elections, the scales on which these two entities believe media matter suggest an enormous chasm. Public perceptions of the power of media in elections, and the academic evidence of its influence, could not be further apart. This essay conveys an understanding of the origins and consequences of this great divide with respect to assessments of campaign media, including both political programming and political advertising.
First, I provide a sketch of how academic thinking on this topic has evolved since the early twentieth century. Second, I explain in greater detail the origins of public beliefs in omnipotent media. I also respond to the counterarguments that are frequently offered up to prove that academics are simply too out of touch with the real world to understand what is actually going on. Finally, I explore the reasons that this gap in understanding has only widened in recent years.
For American citizens, it often seems self-evident that, as the old adage goes, political candidates are sold like soap: they are simply advertised directly to the public.1 In reality, there are fewer similarities than one might expect between the selling of packaged goods and the winning of votes for candidates. Because of these dissimilarities, public assessments of the importance of paid and unpaid media in campaigns may be off by miles rather than inches. Candidates are much more difficult to sell than soap, particularly when they run for high-level offices that attract the most press attention and the strongest claims for media influence. The purpose of this essay is to outline the reasons for this great divide, and then to highlight the seriousness of its consequences for the allocation of political resources.
When academics talk about the effects of mass media on elections, the received history is often described in terms of three distinct periods in scholarly thought about the importance of media in altering mass opinion. This evolution characterizes scholars as initially believing that media had massive effects on political attitudes and opinions, followed by a period in which these effects were assumed to be minimal, and ending with a third era in which such effects were once again assumed to be at least substantial, if somewhat different in nature.
This received view is nothing more than a conveniently reconstructed straw man, with little connection to the weight of scholarly research on media effects at any given point in time.2 In the early part of the twentieth century, to the extent that scholars studied political persuasion at all, they used a case study approach. Between the two world wars, covert propaganda was of particular concern, and many academic case studies were used as part of a large-scale public education effort known as the Institute for Propaganda Analysis. The Institute prepared and distributed instructional materials to schools and adult education groups in order to educate the mass public about how to recognize covert propaganda. The Institute’s reformist mission was to protect the public from potential influence, and the best way to do that, its founders believed, was to heighten public awareness of the threat.3
Historians suggest that one would be hard pressed to find evidence of a scholar from this interwar period claiming actual evidence of the massive effects of propaganda. 4 content However, implicit in the scholars’ meticulous attention to analyzing media messages, and in their desire to protect the public, was the assumption that media at least had the potential for great influence, and thus the public was at risk. Their goal was “to alert the public to the dangers of manipulation.” Many academics from this period “shared an impulse to protect . . . against the new alliance of institutional persuaders and modern communication practitioners.”5 In essence, social scientists engaged in “a kind of clinical social science” in which the public in its entirety was their at-risk patient. Clearly, scholars were worried, but they did not go so far as to claim or document an actual impact from propaganda or from mass media more generally.
From the 1940s onward, researchers have empirically evaluated media influence on political opinions and on vote choice in particular. Sociologist Paul Lazarsfeld and his colleagues at Columbia University initiated this work, using a series of panel surveys of single communities in the United States. These studies, later known collectively as the Columbia Studies,6 suggested that most citizens knew for whom they would vote long before the general election campaign; and in interview after interview, they stuck to that preference. In the original Erie County, Ohio, study of the 1940 election, only 8 percent ever changed their minds between May and the November election. Those few who did change their preference had exceedingly low levels of exposure to political media, thus making it difficult to argue that they were persuaded by campaign communications.7 On this basis, the minimal-effects conclusion was launched.
Starting in the 1950s and continuing to the present, a nationwide data collection effort known as the American National Election Studies (ANES) took over the task of understanding how people decide for whom to vote. However, by the time the ANES was organized and under way, the notion that media had only minimal effects on vote choice was already firmly entrenched in academe. As a result, the ANES directed little effort toward studying media effects, and even if it had done so, the results would most likely have been disappointing. The central pattern originally observed in the Columbia Studies persists today: those most likely to change their vote choice are the least likely to be heavily exposed to political media. Upon reflection, this pattern is not all that surprising; those most heavily interested and involved in politics in this country are also heavily partisan, highly committed to their choices, and thus unlikely to be dissuaded, regardless of any media to which they are exposed. Heavy exposure to media simply does not go hand in hand with a propensity to change opinions.
In fact, now as then, an overwhelming majority of voters decide for whom they will vote many months in advance of an election. Many know that they will vote for the Democratic or Republican candidate even before the parties have officially chosen their nominees. As a result, the number of people who are available as possible targets of persuasion is relatively small. In 2008, for example, between 70 and 80 percent of partisans knew their vote choice well before the general election campaign officially began, thus making it difficult to argue that they were persuaded by any campaign communications. Notably, this proportion is even greater than what Lazarsfeld and his colleagues found in the 1940s. As media scholars point out, presidential elections can be decided by very small margins, so Lazarsfeld’s8 percent who changed preference – or the more recent, substantially lower estimates from 2008 – can still be highly consequential.8 But then as now, few of the changers are exposed to a great deal of political media.
In what is typically characterized as a third era of scholarly study, the consensus has drifted back toward an equilibrium in which most researchers claim evidence of neither massive nor minimal media effects, particularly when speaking to the issue of whether media directly persuade people to support one candidate over another in the context of an election. Although occasional studies demonstrate statistically significant persuasion effects,9 efforts to study entirely different kinds of media influence are now most common. These media effects include learning from media exposure, agenda setting, and priming. As a result of this greater diversity in study outcomes, scholars today often do not define effect in the same way that scholars of these earlier eras did. In particular, the realm of effects of interest has shifted away from media’s direct persuasive influence on public opinion to more subtle and indirect means of altering political processes.
For example, media’s ability to prime certain issues over others has indirect implications for vote choice. To the extent that campaign media emphasize an issue that is perceived as one candidate’s strength or another’s weakness, the voter’s decision-making calculus will be skewed more heavily toward evaluating candidates on that particular issue, which could favor one candidate over another. Few people are single-issue voters, but issue priming could to some extent shift a candidate’s overall favorability. Nonetheless, this process is subtle and indirect relative to more obvious, direct efforts to persuade.
Not everyone has shied away from direct persuasive effects, however. One prominent exception to these generally lower expectations is the perspective advanced by political scientist John Zaller, who argues that media effects are indeed massive on an ongoing basis; we are simply unable to observe them in most observational (that is, non-laboratory) contexts. In short, Zaller denies the absence of evidence as evidence of absence. Instead, he suggests that the gross influence of competitive media in the political environment is huge, but that because the two-candidate and party organizations generally cancel one another out through their persuasive efforts, the net impact of media on opinion is often slim to none. As long as the amount of media is balanced and both sides promote their messages to roughly the same extent and with the same degree of skill, the net influence will appear to be zero even though it results from large amounts of persuasion on both sides. According to this theory, if one side chose, for example, not to advertise, the opponent would experience a landslide victory. But under ordinary circumstances, competing communication flows from each side maintain the status quo.10
This clever idea makes a great deal of sense in many political contexts. It also highlights the need to study situations with large imbalances of media on one side versus the other in order to observe media impact in the real world. Some of Zaller’s work has been able to do just that, primarily in the context of downballot races – for example, elections for the House of Representatives in which one candidate’s communication budget swamps the other candidate’s budget. But such scenarios are still difficult to interpret in unambiguous causal terms. After all, the reason one candidate has so much more money to spend on advertising than the other is typically because he or she is more popular to begin with. Moreover, down-ballot races are precisely the kind in which advertising works most easily. In these cases, persuasion is not necessarily required to change an individual’s vote; name recognition alone may be enough.11
While Zaller’s argument is compelling, it is nonetheless surprising that evidence of media persuasion in politics remains so slim. As a recent review noted, “Volumes of research on electoral communication in recent years have produced precious little evidence of large effects.”12 Although the recognition of new types of effects has meant that scholars now claim at least “not so minimal” influence, current findings are not all that different from the conclusions drawn in the 1970s minimal effects classic, The Unseeing Eye: The Myth of Television Power in National Politics:
Symbolic manipulation through televised political advertising simply does not work. Perhaps the overuse of symbols and stereotypes in product advertising has built up an immunity in the television audience. Perhaps the symbols and postures used in political advertising are such patently ridiculous attempts at manipulation that they appear more ridiculous than reliable. Whatever the precise reasons, television viewers effectively protect themselves from manipulation by staged imagery.13
Others have argued that the lack of evidence in academic research is due to problems in the reliability of media-exposure measures in observational studies. Self-reported survey measures of exposure are indeed suspicious for a number of reasons14; however, their true-score reliabilities are no worse than the kinds of outcome measures they are used to predict.15 Thus, it is difficult to explain why a noisy independent variable is problematic, but a noisy dependent variable is not. Survey research using similarly unreliable measures indeed produces evidence of effects such as political learning. For these reasons, the methodological argument falls short of explaining this pervasive pattern. Yet another possibility is that usual sample sizes do not have sufficient statistical power to detect effects.16
Whether the problem in documenting media influence during campaigns is largely methodological or instead comes from the fact that actual effects are typically much smaller and more infrequent than anticipated when they occur at all, the end result is the same. The small to null effects that can be “teased out of massive electoral communication campaigns” are not terribly impressive.17 Although advertising is just one form of election media, conclusions about the impact of the news are similarly underwhelming, unless one looks for effects other than a change in vote preference, or if one looks at low-profile, local races. The scholarly consensus, specifically on direct persuasive effects of media on vote choice – the type of effect that most fascinates the public and the media – is still that media’s impact is marginal at most. Advertising appears most influential in races for low-level offices where name recognition alone can produce votes. The slick, highly professional advertising that most Americans think of as powerful appears in high-level races such as the presidency – and there is little evidence of direct persuasive effects in these races. Moreover, advertising effects appear to be short lived when they do occur. Although laboratory studies can easily demonstrate what works and what does not, these results are widely believed to be ungeneralizable to the rough-and-tumble world of real politics.
In contrast to the waxing and waning (and rewaxing) of the academic consensus regarding media’s influence on opinions, the American public has consistently believed in very powerful media effects on vote choice and public opinion for a long time. Is this yet another case of poor communication between the academic world and the public? To some extent; but this gap stems more directly from the public’s use of different forms of evidence for inferring media power.
For most Americans, evidence of media’s political power is obvious and omnipresent. After all, they watch television, read newspapers (both online and in print), and see the ads, whether on the air or as a topic of discussion in other media. Thus, foremost among the heuristics that signify media’s power is ubiquity. Media are literally everywhere in Americans’ physical environments. They follow people into their cars, accompany them while on vacation, and permeate day-to-day life. Size matters; in the eyes of citizens, things that are large or widespread are usually also perceived to be important. Indeed, the more publications by a particular author that graduate students are required to read, the taller they will estimate that scholar to be, and the more facial hair they will expect him to have.18
Media’s ubiquity leads people to infer that media must be powerful, if only because its presence dominates all aspects of life and reaches all kinds of people. Although political media flood the airwaves only during election years, Americans think about the sheer number of people reached by these political messages and assume high levels of persuasion from the high visibility of media. Further, many of these messages are obviously designed to be persuasive, so it seems self-evident that they must move opinions.
Americans believe in the political power of television in particular. The 1987 U.S. News & World Report cover heralded “Television’s Blinding Power.” This “telemythology,” as it has since been dubbed by academics, consists of “a set of widely circulated stories about the dangerous powers of television.”19 There is a strong belief among Washington elites “that the general public can be mesmerized by television images. . . . The power of television is perhaps more firmly an article of faith in Washington than anywhere else in the country.”20
In addition to the tremendous reach and visibility of television, most Americans are well aware of the mass persuasion industry and of political consultants and political advertising in particular. Given the received wisdom that politicians are sold just like soap, why shouldn’t the public infer that political ads, like product advertisements, typically persuade people to “purchase” the product? This simple analogy often fails because the political context includes several important differences. First, although there is brand loyalty when one buys soap, it is nothing like the long-term brand loyalty inspired by political parties, which tends to remain stable throughout adulthood. Given that most Americans vote consistent with their preexisting party identification, these persuasive communications are up against fairly powerful adversaries.
Moreover, the product marketplace includes dozens of choices for soap. For this reason, one brand rarely campaigns against another by throwing mud at a specific target. If Dove badmouths Irish Spring, consumers can easily turn to Dial instead of Dove, so negativity is not an efficient approach to boosting sales. Further, as noted above, it is easier to observe effects from product advertising because Dial and Dove seldom launch their advertising campaigns at exactly the same time. When one brand is promoted, but not the other, it is far easier to observe the effectiveness of an advertising campaign in influencing purchases. In the political world, this kind of timing seldom happens; election day is the same for both candidates, so the candidates run their campaign communications more or less simultaneously.
If ubiquity and the analogy to product advertising are not enough to convince Americans of the power of mass media, then surely the prevalence of political consultants will do so. As my nonacademic friends typically argue, “No candidate in his or her right mind would spend that kind of money on something if it didn’t work!” This is an excellent and extremely interesting point. If media are not a powerful force for mass persuasion, why do political candidates spend the bulk of their campaign budgets on media? Indeed, nothing is more visible about the campaign than media consultants:
The airwaves teem with political commercials. The newspapers overflow with commentary about the broadcast spots. And then new TV spots incorporate the print commentary about the old spots. At times candidates and voters seem to be on the sidelines, passively observing the media consultants and ad agencies on the playing field.21
Citizens logically infer that all this activity must somehow make a difference.
To push this argument further, why wouldn’t political media consultants eventually go out of business if they were ineffective at producing the results their candidates desire? The rise of highly professionalized political campaigns is known worldwide as the “Americanization” of campaigns: “The USA is universally acknowledged as the leader in campaign innovation, historically the first to embrace the paraphernalia of political marketing.”22 Campaigns in many other countries have now followed suit because professionalized American campaigns are believed to be more effective.
Professionalized campaigns emerged in the United States not because they demonstrated superior abilities relative to former methods of campaigning but because of the decline in patronage labor to run campaigns. As the patronage system waned in the United States, fewer people volunteered to work for campaigns; thus, commercial firms with paid employees stepped in to fill the void.23Around the same time, the development of computerized voter databases and specialized communications technology encouraged the formation of firms that offer expertise on everything from producing television advertisements, to sending direct mail, to the use of commercial telephone banks.
Does this specialized expertise give a competitive advantage to those who hire campaign consultants? Possibly, but this is far from a foregone conclusion. The limitations of informal observation as a means of assessing effectiveness are severe in the context of elections. As one consultant has noted, the dominant assumption is that “everything you did in a winning campaign was a good idea and everything that you did in a losing campaign was a bad idea.”24 Given that winning or losing is a very crude outcome measure, learning via this kind of evidence takes place only very slowly, if at all. This problem is further complicated by the fact that election outcomes are consistently overdetermined. There are so many factors to which one might attribute a victory (or loss) that one is never certain. Was it the ad campaign? The negative ads in particular? The press’s discussion of the negative ads? The debate performance? A lack of success in getting out the vote on election day? The economy? The weather?25
With as many possible claims as there are components to a campaign, campaign professionals tend to rely on tradition and intuition rather than data. As one seasoned campaign manager noted: “It’s probably the only industry in the world where there’s no market research. . . . But a billion dollars is spent on politics every cycle. No company, no entity, no business would spend that amount of money without knowing what works. . . . No one who gets hired wants to admit they don’t know anything.”26
Indeed, campaign operatives seldom do their own research on what works most efficiently, and they have systematic disincentives to consider independent academic research on these topics if it suggests substantially changing what they do:
Consultants make money by selling specialized expertise (e.g. crafting ads, conducting polls, buying airtime). The profitability of their firms is greatly enhanced by selling the same type of service to a variety of different campaigns. Thus, for example, a given consultant who specializes in running campaigns that rely on direct mail and phone banks has an incentive to manage several campaigns that each rely on these technologies.27
Campaign consultants are heavily invested in certain approaches. They make money by transporting these capacities from place to place and election to election. If what they do is not as effective as has been assumed, they may not want to know about it because that would wreak havoc on their business models: “Few involved in management of campaigns have an interest in developing a clear sense of what works.”28
In a few isolated cases, consultants have collaborated with academics to run scientific field experiments in order to test, for example, which techniques have the greatest effect in increasing turnout.29 But for the most part, consultants are uninterested in empirically validated best practices and prefer to stick with folk wisdom. Tracking polls, which show overtime trends in a candidate’s standings, are about as close as they come to gathering evidence that allows them to ascertain whether one approach works better than another. But in an uncontrolled campaign environment in which everyone receives the “treatment,” there are typically so many potential interpretations of what caused any observed change that strong causal inference is impossible.
Often, the knowledge gained can only benefit those campaigns that follow the one invested in the research. As one campaign operative complained, “Finding out the day after the election that Treatment A was the best is of limited value to an organization like ours. We’re actually trying to win the election.”30 Moreover, if you are a campaign consultant, trying something new can easily lead to blame for a loss, whereas sticking with what everyone else does carries less risk. Given that there are no independent firms systematically monitoring the effectiveness of campaign strategies, sticking with what is assumed to matter most is the safest strategy.31
If campaign professionals generally do not execute these kinds of studies, then why not academics? Do political scientists or market researchers have useful things to say about campaign tactics? Neither the National Science Foundation nor any other foundations fund research on “what works” to gain votes for one candidate over another; this would be considered partisan spending. The closest academics come to this kind of focus is research on turnout. Because turnout is considered a public, nonpartisan good, research on this topic is widely funded in academe. Certainly, knowledge about how to increase turnout can be used for partisan purposes when areas favorable toward a given candidate are targeted for increased turnout while others are not. But studies of political persuasion in the context of campaigns are seldom the focus of academic research because of their partisan implications. As a former editor of Campaigns and Elections suggests, “Practitioners think that political scientists are not studying problems of interest and are therefore not helpful.”32
To summarize, a combination of factors collude to make elusive any well-controlled empirical research on how media can most efficiently influence public opinion. First, the accumulation of knowledge is hindered by the fact that campaign consultants are reluctant to participate in the research that would be necessary to find out how to use media most efficiently. They fear that purposely not exposing parts of the population to their media will lose votes. When one campaign did sign on for an experimental field study, the move “potentially set one campaign manager up for malpractice.”33 No one wants to undermine the chance for victory. The fact that only subsequent campaigns might benefit from the research provides another disincentive. Finally, campaign consultants’ business models rest on certain assumptions that, if untrue, could prove financially disastrous for them.
Ironically, America leads the world in spending huge amounts of money on something that only possibly accomplishes what it sets out to do. If campaign media does persuade voters, it does so highly inefficiently. In reality, “The claim of political savants and insiders that the right commercials and the right consultants can win any election . . . is fed by the self-serving myth that certain ‘magic moments’ on television have turned elections around.”34
But perhaps more important, because observers of campaigns perceive highly professionalized campaigns to be more likely to succeed, candidates continue to pay huge amounts to campaign professionals, who continue to rely on instinct and tradition in spending candidates’ money. There is a self-fulfilling aspect to the professionalization of campaigns. If a candidate does not spend large sums on television (the least efficient of campaign communications), then he or she is seen as less “serious” as a candidate. This impression can impair fundraising ability and the candidate’s perceived electability, even if the ads themselves affect no one.
The visibility and professionalism of campaign media heavily influence perceptions of its potential impact:
In an environment where very little is known about what kinds of campaign tactics actually work, those who purchase these campaign services must rely on their intuitive sense of what makes for an effective campaign. There is a natural tendency to gravitate toward tactics that command the attention of others, particularly potential donors. Campaigns crave attention and credibility: expensive, large-scale, professionally crafted communication is a way to demonstrate one’s seriousness of purpose.35
Thus, while the general public associates greater professionalism with greater impact, research findings often suggest otherwise. Expensive television advertisements attract a great deal of attention, but they may be one of the least cost-effective means of persuading voters.
A psychological tendency known as the persuasive press inference, or third person effect, further exacerbates the public tendency to perceive large media effects. More educated and involved partisans are especially likely to perceive that others are influenced by media, though certainly not themselves. Their assessments of the extent of influence from any given message systematically exaggerate the amount of influence. As a result, “The power of the media resides in the perception of experts and decision makers that the general public is influenced by the mass media, not in the direct influence of the mass media on the general public. That is to say, the media’s political appeal lies less in its ability to bend minds than in its ability to convince elites that the popular mind can be bent.”36
Today, the great divide between public and academic perceptions of media influence on vote choice may be widening still further. The more overtly partisan political media environment has led many academics to assume that the potential for changing preferences through political news has waned a great deal. As news and talk shows become more plentiful and increasingly partisan, citizens can more easily self-select like-minded programming that is unlikely to change their preferences as much as reinforce them. To what extent these theories of waning influence will be borne out has yet to be observed, but many scholars have speculated that individuals’ exposure to ideas they do not already agree with will be increasingly limited, thus making persuasion unlikely as well. Thus, academics have already begun to note “the waning of mass media influence in the lives of most citizens.”37
The mass public, on the other hand, looks at some of the programming on offer today and finds it to be heavily biased toward one candidate or the other – more so than in the past. As a result, the public sees the potential for persuasive influence from media as greater than ever before. Without taking into account the likely audiences for these programs, the content itself seems far more hard-hitting and potentially persuasive than the news programs of the past, which at least attempted to achieve balance and neutrality.
Further, through a bizarre trend dubbed media narcissism, self-reflexive reporting, or metacoverage, media have become fascinated with themselves as a political force, and they increasingly cover their own importance in the political process as a standard part of election coverage.38 According to many journalists, the campaign story has become the analysis of candidates’ use of media to manipulate the public into voting for them. In 1980, one reporter claimed, “Never before, it seemed, had so many reporters, correspondents, editors, executives, candidates, consultants, and just plain citizens been so conscious of the power of the press.”39
It is doubtful that this phenomenon emerged full blown in 1980; after all, popular assumptions about the importance of media in winning elections were also high in the 1940s. According to popular legend, Roosevelt’s victories were attributable to his “superb radio voice,” which enabled him to exploit the medium better than Landon or Willkie.40 And books such as author Joe McGinniss’s The Selling of the President offered entertaining anecdotal tales of media power to popular audiences in the 1960s. But according to scholars, media metacoverage has increased. Thus, in 1988, Michael Dukakis’s loss was attributed to his failure to “package” himself successfully for the demands of media politics, and George H.W. Bush’s success was attributed to his superior media consultants.41 The amount of time the media spend talking to and about themselves has increased relative to the time they spend talking about actual politics. In short, the media have shifted their focus increasingly to themselves.42
Metacoverage is obviously self-serving to a degree, in that media are continuously celebrating their own importance in the political process. But my view is that this practice stems from more than a sense of self-importance. Rather, political media see themselves as a modern-day Institute for Propaganda Analysis, focusing on strategy and tactics in an effort to prevent the unwitting public from becoming victims of political persuasion. By covering potentially persuasive media, they make us “cognoscenti of our own bamboozlement”43; they make us feel as if we are smarter than others who may fall prey to these tactics.44 The same protective impulse that drove early assessments of propaganda’s potential still influences journalists’ perceptions of their audiences’ susceptibility. Moreover, expressing cynicism about persuasive appeals makes the individual cynic feel smarter, and media coverage of politics encourages us to be among those “in the know.” Although the press’s intentions may be good, this portrayal of campaigns and elections is not a flattering portrait of the public or of the political process. The public is viewed as gullible and easily manipulated by all things nonpolitical, and the process itself is portrayed as a sham.
The extremely high levels of spending on American political campaigns are a perennial cause for disdain of the American electoral process. For the 2012 presidential election, the Supreme Court’s 2010 Citizens United v. Federal Election Commission decision has become the whipping boy, but other previous decisions, such as Buckley v. Valeo in 1976, have produced similar outcries about the increasingly high costs of elections. The underlying reason that people are upset about the amount spent on campaigns is that they believe money buys television airtime, which, in turn, buys votes. When television time does buy votes, it does so highly inefficiently. Thus, my own complaint is somewhat different: the problem with the high costs of campaigns is that such huge amounts of money are spent unproductively and inefficiently when they could be spent in ways that more directly affect Americans. Despite the rise of narrowcasting, television is still among the least efficient means of persuasion, dollar for dollar. But the high costs of television and its perceived necessity mean that political leaders feel they must spend more and more of their time raising money rather than governing.
For a variety of reasons, media influence is indeed a difficult topic to study outside the laboratory. But regardless of the extent to which media actually influence election outcomes, we are not, as a political culture, well served by these extreme beliefs in media power. My problem with this common approach to covering campaigns runs deeper than the usual gripe, which is that coverage of strategy and tactics displaces more serious coverage of the campaign. The real problem stems from our culture’s underlying attitude toward political persuasion more generally.
I was struck by this underlying assumption when the human subjects committee at one of my former universities decided that political persuasion was a form of harm to human subjects. Even though the experiment involved nothing more than exposing subjects to highly substantive political arguments on different sides of an issue, this approach was deemed potentially harmful to research participants because their political views might be changed in all manner of directions. The Institutional Review Board wanted the study participants to be persuaded back to their former opinions at the conclusion of the study. This struck me as absurd. If persuasion equals harm, then our political system has some pretty serious problems on its hands. The entire purpose of election campaigns is to provide politicians with opportunities to expose the public to their persuasive arguments. Persuasion, rather than coercion or violence, was thought by our Founders to be a preferable means of conducting politics. But today we are ambivalent, at best, about this core part of our political system.
Presidential candidates spend around 70 percent of their extremely large campaign budgets producing and airing political ads.45 Even in the relatively low-profile mid-term elections in 2010, candidates spent around $1.5 billion. Outside groups, such as Super PACs, now add substantially to total campaign spending. Professionalization of campaigns sends a signal to citizens that these people know what they are doing; they have expertise that we do not, so we are the potential victims of their efforts to manipulate us. But if the emperor has no clothes, then what?
The real tragedy here is not that so much is being spent or that people are being persuaded to change their minds willy-nilly, but rather that so much is being spent without effect. If campaigns effectively inspire, galvanize, and mobilize the American public, it is easier to defend their massive expenditures on media. But if they do so only through a highly inefficient waste of time and resources, then this reality is indeed regrettable.
Does it matter that the mass public believes in large media effects instead of smaller, more realistic ones? Beyond sheer waste and inefficiency, the tremendous emphasis of journalists on media power in elections, and the corresponding strong belief in its influence among the mass public, have negative consequences for the perceived legitimacy of our system of government. It has long been argued that political participation in the form of voting in elections is an especially inefficient way to express one’s political views. The chance that one’s vote will matter is infinitesimal, and even if one is lucky enough to have supported the candidate who ultimately wins, our political system is sufficiently complex that there is no guarantee that the elected official will be able to accomplish his or her specified goals. What makes elections legitimate in the eyes of the mass public is not that the electorate always gets its way. Instead, the process itself is what confers legitimacy on the outcome. But if the process is believed to be a function of who hired the better political consultants or who spent more on advertising, then it becomes very difficult for those on the losing side to see the election outcome as legitimate.
If elections are believed to be won and lost because of the tactics of professional campaign consultants – not because of the beliefs of the mass public, or the merits of candidates, or politics – then how can the outcome be respected? As one observer put it, “What better excuse than that the game was rigged, the press bought, the television networks intimidated . . . and voters led like lambs to the polling booths.”46 Today, there are strong differences of opinion among Americans about the appropriate role of government in society and about how that government should be run. These are real differences, not made-for-TV conflicts. Sadly, the “mythology of the great power of U.S. election campaign practices” does little to advance public understanding of or respect for these very real differences.47
Finally, in addition to wasted resources and less perceived legitimacy in election outcomes, beliefs in the power of campaign media ultimately elevate media’s actual power in elections. It is a cliché to say that politics is about perceptions, but it is also true. As long as 90 percent of the American public believes that the news media influence who becomes president48 and more than 70 percent see that influence as growing,49 candidates and their campaigns will continue to behave as if these perceptions were true. To do anything else risks being seen as less serious and, therefore, less electable.
1 Alex Marland, “Marketing Political Soap: A Political Marketing View of Selling Candidates Like Soap, of Electioneering as a Ritual, and of Electoral Military Analogies,” Journal of Public Affairs 3 (2) (May 2003): 103–115.
6 The first two of these studies focused on the 1940 and 1948 elections, respectively, and were published as Paul Lazarsfeld, Bernard Berelson, and Hazel Gaudet, The People’s Choice: How the Voter Makes Up His Mind in a Presidential Campaign (New York: Duell, Sloan and Pearce, 1944); and Bernard Berelson, Paul Lazarsfeld, and William McPhee, Voting: A Study of Opinion Formation in a Presidential Election (Chicago: University of Chicago Press, 1954).
7 For a review of the Columbia Studies’ findings with respect to media influence, see Steven H. Chaffee and John L. Hockheimer, “The Beginnings of Political Communication Research in the United States: Origins of the ‘Limited Effects’ Model,” in The Media Revolution in America and Western Europe, ed. Everett M. Rogers and Francis Balle (Norwood, N.J.: Ablex, 1982), 263–283.
9 See Jonathan McDonald Ladd and Gabriel S. Lenz, “Exploiting a Rare Communication Shift to Document the Persuasive Power of the News Media,” American Journal of Political Science 53 (2) (April 2009): 394–410.
10 John Zaller, “The Myth of Massive Media Impact Revived: New Support for a Discredited Idea,” in Political Persuasion and Attitude Change, ed. Diana C. Mutz, Richard A. Brody, and Paul M. Sniderman (Ann Arbor: University of Michigan Press, 1996), 17–79.
11 See Shanto Iyengar and Jennifer McGrady, Media Politics: A Citizen’s Guide (New York: W.W. Norton, 2007); Michael L. Rothschild and Michael L. Ray, “Involvement and Political Advertising Effect: An Exploratory Experiment,” Communication Research 1 (3) (July 1974): 264–284; Gary C. Jacobson, The Politics of Congressional Elections, 7th ed. (New York: Pearson/Longman, 2009).
18 Thankfully, this pattern does not hold for female academics, although the data in this case suffer from a small sample size. Data for this unpublished study were gathered from first year Ph.D. students at Stanford University.
38 Matthew R. Kerbel, Remote and Controlled: Media Politics in a Cynical Age (Boulder, Colo.: Westview Press, 1995); David Stebenne, “Media Coverage of American Presidential Elections: A Historical Perspective,” in The Finish Line: Covering the Campaign’s Final Days, ed. Everette E. Dennis (New York: Freedom Forum Media Studies Center, 1993); Frank Esser and Paul D’Angelo, “Framing the Press and the Publicity Process: A Content Analysis of Meta-Coverage on Campaign 2000 Network News,” American Behavioral Scientist 46 (5) (2002): 617–641.
42 Robert S. Lichter, Richard E. Noyes, and Lynda Lee Kaid, “No News or Negative News: How the Networks Nixed the ’96 Campaign,” in The Electronic Election: Perspectives on the 1996 Campaign Communication, ed. Lynda Lee Kaid and Diane G. Bystrom (Mahwah, N.J.: Erlbaum, 2000), 3–13.
48 This figure is based on responses to the following question: “Thinking more about the news media’s coverage of presidential campaigns, please tell me whether you strongly agree, somewhat agree, somewhat disagree, or strongly disagree with each of the following statements. . . . Media coverage has too much influence on who Americans vote for”; Confidence in Leadership Survey, September 2007. Data are provided by the Roper Center for Public Opinion Research, University of Connecticut.
49 This figure is based on responses to the following question: “Please tell me whether you think the news media today have more, less or about the same influence as they did 40 or 50 years ago [on] [w]ho becomes President”; Roper Starch Worldwide, January 27–30, 2000. Data are provided by the Roper Center for Public Opinion Research, University of Connecticut.