Oxford Handbooks Online

Oxford Handbooks Online Campaigning, Debating, Advertising Bradford H. Bishop and D. Sunshine Hillygus The Oxford Handbook of American Public Opinion and the Media Edited by George C. Edwards III, Lawrence R. Jacobs, and Robert Y. Shapiro Print Publication Date: May 2011 Online Publication Date: Sep 2011 Subject: Political Science, U.S. Politics, Political Behavior DOI: 10.1093/oxfordhb/9780199545636.003.0013 Abstract and Keywords This chapter studies debating, advertising, and campaigning, the latter playing an important role in American democracy, first discussing the development of political campaigns and summarizing the broad theoretical perspectives that have directed scholarly thinking and research on campaign effects. Next, it examines the findings of modern research on the effects of the different sources of campaign information, along with their implications for the democratic process. The chapter ends with a discussion of the opportunities and challenges for future research. Keywords: political campaigns, campaign effects, campaign information, democratic process, advertising CAMPAIGNS play a critical role in American democracy. They are a forum for public debate that allows for learning, comparing, and deliberating about the major alternatives competing for elected office. The effects of political campaigns are complex and multifaceted, a dynamic interaction among many players who influence one another. Together, they determine the outcome of the electoral process and the meaning of that outcome. The study of campaign effects has evolved in fits and starts since the turn of the twentieth century with a flurry of research in recent decades offering compelling evidence that campaigns can and do have a measurable influence on the beliefs, attitudes, and behaviors of the voting public. The purpose of this chapter is to critically examine and synthesize this field of research and to consider possible directions for future research. The chapter proceeds as follows. First, we discuss the development of the field, outlining broad theoretical perspectives that have guided scholarly thinking and research on campaign effects. We then consider the findings of contemporary research regarding the effects of various sources of campaign information and their implications for the democratic process. Finally, we consider some of the challenges and opportunities for future research. Development of a Field Students of political campaigns are familiar with the uneven development of the field; the traditional starting point for most research on campaign effects is a discussion of (p. 205) the prevailing scholarly perspective of minimal effects. Not only has recent research effectively reversed this once conventional wisdom, the minimal effects thesis was always a bit of a straw man the consequence of a particular conception of campaign effects and a particular interpretation of early research. This perspective is traced to the classic Columbia School studies of the 1940s, in which Lazarsfeld and his colleagues found that few people changed their minds during the campaign, and those who did change their minds typically did so in a manner consistent with their predispositions (Lazarsfeld, Berelson, and Gaudet 1948; Berelson, Lazarsfeld, and McPhee 1954). Since the majority of respondents (71 percent) settled on a preferred candidate before the start of the fall campaign, it was reasoned that campaigns had only a minimal impact on voting behavior. Page 1 of 12

This perspective was itself a reaction to the previous scholarly thinking about propaganda effects, the so called hypodermic needle theory, whereby mass media were thought to have a direct and powerful effect on a passive and uninformed public. This expectation about massive media effects was advanced by critical theorists who attributed the rise of the Nazi Party in Germany to the use of propaganda messages in radio and film (Lasswell 1927). In the United States this perspective seemed to be validated by the 1938 Orson Welles radio production of The War of the Worlds, which many listeners believed were actual newscasts about an alien invasion. Given the expectations and worries about the possible deleterious effects of propagandistic persuasion, the minimal effects conclusions of the Columbia School studies seemed both reassuring and prudent. The minimal effects perspective was reinforced by later political science research finding a correlation between voting behavior and long term psychological characteristics. In particular, the perceptual screen of party identification was hypothesized to shield people from persuasive campaign rhetoric (A. Campbell, Converse, Miller, and Stokes 1960; Converse 1964), offering an explanation for voter decision-making in the absence of political knowledge. Subsequent research challenged some of the findings of the early voting research, but did not universally place campaigns at the center of the story. For example, V. O. Key (1966) presented an electorate that based vote decisions on more coherent, substantive criteria, but he argued that retrospective assessments of incumbent party performance held greater sway than campaign rhetoric: Campaigning does change votes and it does bestir people to vote. Yet other influences doubtless outweigh the campaign in the determination of the vote. As voters mark their ballots they may have in their minds impressions of the last TV political spectacular of the campaign, but more important, they have in their minds recollections of their experiences of the last four years. (Key 1966, 9) Building on the logic of retrospective voting, forecasting models were able to predict election outcomes using just a few fundamental variables such as presidential approval and economic conditions (for example, Rosenstone 1983). Overall, then, political science research did not argue that campaigns completely failed at inducing supporters of one candidate to support another candidate: these effects were simply confined to the margins. In the absence of large, visible (p. 206) shifts in citizen voting intention during the campaign, scholars were reluctant to abandon the academic conventional wisdom (Holbrook 1996, 10). And even the shifts in vote choice that were observed were thought to be largely predictable (Gelman and King 1993; J. Campbell 2000). Finkel (1993) argued that campaigns had the potential to be determinative, but rarely were in presidential elections because both parties employ highly competent and skilled campaign professionals; campaign efforts thus counterbalanced one another, eliminating the possibility that a single campaign strategy or effort could be persuasive enough to shift outcomes entirely in one direction. By the 1990s, however, there was renewed interest in campaign effects (Holbrook 1996; Shaw 1999), fueled in part by increased campaign spending, by declining party identification, and by the availability of more and better data that allowed scholars to more closely link election efforts with voter behavior. This literature has accumulated a growing body of evidence that campaigns do shape voting behavior and election outcomes. In addition to better data and more sophisticated methods, a notable reason for the shifting conventional wisdom is a simple change in what is deemed a campaign effect. While the early scholars were looking for evidence of massive shifts in public opinion arising from the pronouncements of charismatic politicians, subsequent researchers sought to measure narrower, more carefully defined, campaign effects. It was clearly not the case then or now that skillful campaigners were persuading vast swaths of the electorate to reverse their policy positions and candidate preferences. Although voters may not be sophisticated ideologues, they are not tabulae rasae; they bring prior beliefs, affiliations, and interests with them to the context of the presidential campaign and these characteristics influence how they react to various campaign efforts. Because some individuals may be more or less receptive to campaign efforts, it can often require large data sets in order to capture evidence of campaign effects among small subgroups in the population who are susceptible. Second, it may have been too blunt to try to measure persuasion simply by looking at swings in candidate choice. It is worthwhile to understand voters' shifts from undecided to a candidate selection, for instance. More generally, if we think about voters as carrying a latent probability of supporting one candidate or the other, it is not clear that the only significant effects are those that move someone over the threshold of.50 in a two candidate race. We can Page 2 of 12

imagine cases in which those nearest.50 may be moved from one candidate to the other with only the slightest stimulus, while those near the ends of the distribution might experience significant campaign effects that result in large shifts in their latent probability, but nonetheless do not push them over the prediction threshold. Today, scholars also recognize that campaigns may influence the public in a variety of indirect ways. Campaigns can increase the amount of information citizens hold about the candidates they are asked to choose between: a learning effect. Campaigns can reconfigure the mix and relative weights of considerations people rely upon when arriving at a vote choice: a priming effect. Campaigns can induce some citizens to go to the polls who otherwise would not have voted: a mobilization effect. (p. 207) It is worth noting that such indirect effects have long been documented, including within the seminal minimal effects literature. Lazarsfeld, Berelson, and Gaudet (1948) readily acknowledged the presence of priming effects: changes to the type and salience of various considerations used by voters when evaluating candidates. Berelson, Lazarsfeld, and McPhee (1954) also observed voters changing voting intention from undecided to favoring a particular candidate. Kramer (1970) offered evidence of a mobilization effect, concluding that campaigns do not change the preferences of perennial voters but are able to convert prior non voters into partisan supporters. And Nie, Verba, and Petrocik (1976) argued that the public's capacity to engage in issue voting was in large part conditioned by whether politicians offer meaningfully different positions on those issues, presuming a learning effect. In short, contemporary characterizations of the minimal effects perspective often oversimplify conclusions of the classic research (Shaw 2006). Another evolution in campaign research has been the debate and discussion regarding the normative implications of political campaigns. Normative concerns about the malleability of an uninformed public animated early campaign research, but were largely refuted by the empirical evidence. There is little evidence that campaigns manipulate individuals to vote against their self interests; rather, they appear to help voters connect their often complex and competing predispositions with a candidate selection (Hillygus and Shields 2008). Thus, whereas campaign effects were once viewed as detrimental to the health of democracy, they are now largely viewed as contributing to successful functioning of the electoral process (for example, Freedman, Franz, and Goldstein 2004). Even still, broad normative questions remain. Do campaigns increase deliberation about the policies, traits, and backgrounds of the candidates or do they simply increase psychological links to the political parties? Do campaign efforts engage and mobilize voters, or can they also serve to alienate them? Do campaigns narrow or widen the wellknown gaps in political awareness and participation among citizens at higher and lower levels of socioeconomic status? In evaluating the contemporary empirical literature, these normative questions should continue to motivate analyses and debate in the field. What Matters? Defining the Campaign While there is now broad recognition that the effects of campaigns are varied, there is no consensus about how the campaign itself should be defined, operationalized, and measured. Some research has focused on discrete campaign events such as debates, conventions, or other consequential occurrences that arise during the campaign. Other literature decomposes the campaign into its constitutive elements like television advertising or candidate endorsements. Still others attempt to proxy campaign information and activities with measures of time or geography, assuming that people will have been exposed to more campaign information by the end of the campaign than at the (p. 208) beginning or that campaigns are more salient in a battleground state than in a safe state. In summarizing the state of literature here, we organize it based on the source of that information, whether campaign advertising, media coverage, or direct communication. There may be alternative and perhaps even more intuitive ways of dividing the research (by the type of campaign effect, for instance). Certainly some of the strongest work of the past decade looks across a multitude of actions, activities, and events (for example, Shaw 1999, 2006; Gerber and Green 2000a). But our approach highlights many of the important gaps in our knowledge and calls attention to the fact that this literature has not always evolved as a coherent and iterative line of research. Campaign Advertising Much of the recent campaign research has focused on television advertising, and for good reason: political Page 3 of 12

advertising takes up the majority of a candidate's budget and it provides candidates with an unmediated method of communicating with voters. Perhaps because of this, there is a large body of research examining the role and impact of advertising and documenting learning, priming, persuasive, and mobilization effects (for a great review, see Franz, Freedman, Goldstein, and Ridout 2007). Despite the high incidence of factual misstatements in television ads, even short television spots have been found to increase the public's knowledge of the candidates and their issue positions (Johnston, Hagen, and Jamieson 2004). Freedman, Franz, and Goldstein (2004) found that citizens who are exposed to campaign ads are more attentive to the campaign, more aware of who is running, and have more to say about the candidates, particularly at lower levels of political knowledge. And such learning has consequences for voter decision-making, increasing the likelihood of issue voting (Brians and Wattenberg 1996). Unresolved, however, is the question of who benefits most from campaign learning. Some research finds that learning effects are concentrated among the most sophisticated, and thus campaigns increase the knowledge gap in the electorate (for example, Holbrook 2002). Others contend that campaign learning primarily affects the uninformed, narrowing differences in knowledge levels between citizens (for example, Freedman, Franz, and Goldstein 2004). Still to be determined, then, are the specific conditions under which political advertising can in fact democratize knowledge in the broader electorate. Scholars have also found evidence that television advertising helps to make particular issues accessible to voters (Johnston, Blais, Brady, and Crete 1992). Though the evidence of priming is robust in the campaign literature, researchers have not yet resolved when and why some considerations may be primed over others. For instance, some argue that partisanship is most likely to be activated by the campaign (Berelson, Lazarsfeld, and McPhee 1954; Kramer 1970; Holbrook and McClurg 2005), while others argue ideological and issue based considerations are primed at the expense of partisan attachments (Gelman and King 1993; Bartels 2006; Hillygus and Shields 2008). This debate, too, has clear normative implications since it could be argued that some (p. 209) considerations could be better than others. For example, campaigns have been found to increase the relationship between racial antipathy and vote choice by invoking racial code words (Mendelberg 2001). In addition to learning and priming effects, scholars have found that campaign advertising is associated with persuasive effects not surprisingly, since most ads explicitly encourage citizens to vote for a particular candidate. Research has found that a candidate's television spending in a state directly translates to an increased vote share (Shaw 1999; Johnston, Hagen, and Jamieson 2004). Other research finds individual level evidence that exposure to television advertising has a persuasive influence on voter decision-making (Huber and Arceneaux 2007; Freedman, Franz, and Goldstein 2004). And making it one of the rare cases where observational and experimental research reach the same conclusion, laboratory findings have also shown that televised campaign advertisements can induce a voter to support one candidate over another (Iyengar and Simon 2000). Another way that campaign advertising is hypothesized to influence election outcomes is by inspiring people who would not have voted otherwise to go through the trouble of getting to the polls (Freedman, Franz, and Goldstein 2004). Nonetheless, there remains some debate in the literature about the magnitude of the mobilizing effect, who is mobilized by campaign advertising, and what kinds of ads are most effective at inducing citizens to vote (for example, Holbrook and McClurg 2005; Krasno and Green 2008). At the same time, there are also questions about whether campaign advertising especially negative advertising might actually demobilize voters. Ansolabehere and Iyengar (1995) argue that negative advertising lowers overall levels of trust in government or enthusiasm for the democratic process. Kahn and Kenney (1999) similarly conclude that campaigns involving non policy personal attacks and mudslinging were associated with lower levels of turnout, while policy based negative advertising was not associated with demobilization, a result that suggests voters distinguish negative personal attacks from policy based ones. On the other hand, considerable subsequent research has challenged the demobilization hypothesis, with some finding that negative ads actually raise turnout indirectly by increasing the amount of information and interest in the campaign among citizens (Geer and Lau 2006) and others finding no impact on turnout (Clinton and Lapinski 2004). If there is consensus on negative campaigning in general, it seems to be that negative ads do promote some degree of learning (Ansolabehere and Iyengar 1995; Lau, Sigelman, and Rovner 2007). Lau and Rovner (2009) provide an excellent review of the lengthy body of research investigating this controversial topic. Although the television advertising research has shown clear evidence of campaign effects, this research highlights the many opportunities for additional scholarly learning. 1 For instance, scholars have not yet settled Page 4 of 12

upon a consensus view about the stability and durability of attitude changes emerging from exposure to TV spots. The observational research also tends to focus on the volume of advertising the number of advertisements in the media market or state. While there are exceptions (the extensive (p. 210) research on tone, for instance), it is much less common to carefully consider the influence of the content of television advertising issue and image content, fear vs. humor appeal, and so on. At the cutting edge of such a focus is recent research by Brader (2005) that shows advertising effects are conditional on emotional content. Telephone, Direct Mail, and Personal Contact While the air war communications continue to receive the bulk of campaign money in most political campaigns, recent campaigns are giving renewed attention to ground war efforts, including direct mail, telephone calls, and personal canvassing. In recent years, there has been a decline in the efficiency and effectiveness of television advertising; with the expansion of media choice (on television and the Internet) and the widespread use of technology like remote controls and TiVo, it has become easier for the public to avoid political advertising (Prior 2007). Candidates now face a higher cost per impression to reach a voter with a broadcast message, so they have turned to email, direct mail, phone calls, and personal canvassing to narrowcast campaign messages to targeted audiences (Monson 2004). Historically, research on ground war activity has focused on grassroots mobilization efforts the targeting of core supporters to get them to the polls. One of the consistent findings in the mobilization literature is that personal contact plays a consistent role in getting out the vote (Rosenstone and Hansen 1993; Verba, Schlozman, and Brady 1995). Lower levels of voting, according to this argument, were a consequence of a less personal, increasingly mediated form of contact between parties and voters. Although an influential body of research, much of the work on party contact is also susceptible to the criticism that the observed relationship between exposure and various political outcomes is spurious, the result of campaign targeting rather than campaign effects. Field experiments have been used as a way to address these concerns, and they come as close as possible to isolating the causal effects of personal contact. Gerber and Green's (2000a) field experiment found that face toface canvassing increased turnout by more than 5 percent, a finding that has been replicated in subsequent studies (Green, Gerber, and Nickerson 2003). Looking at direct mail, Gerber and Green (2000b) found that nonpartisan get out the vote (GOTV) postcards increased turnout among Independents by 7 percent, but had no effect on Democrats or Republicans. Increasingly, candidates are also contacting voters directly through the Internet and email, but the effects of these sources of information are difficult to study, especially because voters tend to select exposure (Bimber and Davis 2002). In one of the few studies on the topic, Nickerson (2007) found that non partisan email contact does not increase registration rates or lead to higher turnout, but clearly more research on the topic is needed. Future research on the ground campaign should focus more closely on the persuasive impact of direct contact. Most of the research in the area has tended to focus on non partisan voter mobilization efforts (for example, Green and Gerber 2004), but the (p. 211) reality of presidential campaign activities is that very few communications are pure GOTV messages. When a campaign sends direct mail, makes a phone call, or knocks on a door, they don't simply ask the recipient to go to the polls without also telling them who to vote for or why they should vote that way a clear attempt to persuade. In an analysis of county level voter behavior, for instance, Masket (2009) finds that counties that had an Obama field office in 2008 had a sharper increase in Democratic vote share relative to 2004. Future research would also benefit from studying who is responsive to which messages. In a meta analysis, for instance, Arceneaux and Nickerson (2009) find that GOTV efforts are most effective at increasing the turnout likelihood of low propensity voters in high intensity elections and of high propensity voters in low intensity elections. Campaign Visits One of the basic campaign activities in presidential elections is the simple decision by candidates to make an appearance during the campaign. Althaus, Nardulli, and Shaw (2002) found that candidates are increasing the number of appearances they make, and they are appearing in out of the way places they may have avoided in years past. In an analysis of presidential campaigns between 1988 and 1996, Shaw (1999) found that candidate Page 5 of 12

appearances were positively correlated with statewide vote share, particularly in states with large percentages of undecided voters. Subsequent pooled time series cross sectional analysis by the same author for the 2000 and 2004 presidential elections concluded certain campaign appearances were associated with increased candidate favorability ratings, and changes in vote choice (Shaw 2006). Less clear is the mechanism by which candidate visits have an effect, whether through increased media coverage, direct contact with voters, or another factor. Campaign Events A number of studies have also found evidence that specific campaign events can influence public opinion and behavior. Holbrook (1996) catalogued major incidents in US presidential campaigns between 1984 and 1992 and found that the relative standing of the two major party candidates is impacted by these events. Scholars have observed that party conventions provide a bump for the convention candidate (Holbrook 1996; Hillygus and Jackman 2003). Campbell, Cherry, and Wink (1992) estimate the convention bump is 5 to 7 points and persists into the general election campaign. Debates similarly offer the public unmediated exposure to the major party presidential candidates' views and personality characteristics, at least until the talking heads begin to contribute their interpretations of the outcome. But their effects on the public may be limited by the tendency of well prepared candidates to offer canned responses (p. 212) to questions and to avoid direct confrontation (Schrott 1990); presidential debates may be more like dual press conferences than an actual debate. As a consequence, we might expect the candidates' messages either to cancel one another out or simply to reinforce preexisting opinions. Certainly, debates offer evidence that the public engages in biased processing (Chaiken, Liberman, and Eagly 1989), where prior attitudes shape how people confront new information (Lanoue and Schrott 1991). Nonetheless, some scholars have found the debates do have an impact on citizens' perceptions, though the mass media again play an intervening role (Holbrook 1996; Johnston, Blais, Brady, and Crete 1992). For instance, Johnston, Hagen, and Jamieson (2004) found that debate viewers perceived Al Gore to be the winner of the first and third presidential debates, though media criticism of Gore's infamous sighs and grimaces led debate non viewers to perceive that Bush had won. A long series of studies also finds that debates increase viewers' knowledge of candidates' issue positions (Lanoue and Schrott 1991; but see Lanoue 1995), though there are conflicting findings about whether debates increase knowledge primarily among low information viewers or increase the knowledge gap between different classes of viewers (Bartels 1993). Holbrook (1999) found the debates do facilitate learning, and the first debate is particularly effective in this regard. Thus, there is some evidence that the debates increase the overall level of knowledge in the electorate, though it is not clear from a normative point of view whether debates level the informational playing field among citizens from different social strata. While the above events are known to competing candidates and their strategists well in advance, other events arise suddenly and unexpectedly, and can have consequences in the election the so called October surprise. For example, some scholars have argued that the economic collapse was decisive in the 2008 election (Linn, Moody, and Asper 2009; but see Hillygus and Henderson 2010). Such research is in its relative infancy in large part because scholars have only recently had access to the data necessary for such analysis. New and emerging data collection efforts, such as the Annenberg rolling cross sections (Johnston, Hagen, and Jamieson 2004) have increasingly made it possible to isolate the effects of specific campaign events, even those unexpected and unscheduled. Mediated Messages To this point, we have focused on the consequences of active efforts by campaigns to influence election outcomes; however, many voters are influenced by information not produced by the campaigns. Indeed, since the public expects messages provided by the candidates or parties to be biased, source credibility and thus the persuasive power of other information sources might well be greater. For instance, scholars have found that interpersonal discussions with friends, family, co workers have a large influence on candidate evaluations (for example, Beck, Dalton, Greene, and Huckfeldt 2002). Huckfeldt and Sprague (1995) report that people seek out informed and (p. 213) knowledgeable people when they seek to communicate about politics, and that people's attitudes are influenced by these contacts. These informal conversations have also been shown to have consequences for subsequent attitude formation and political behavior. Druckman and Nelson (2003) found that Page 6 of 12

interpersonal exposure to diverging points of view can eliminate elite framing effects, while Mutz (2006) finds it can reduce the likelihood of voting. Likewise, many scholars regard media coverage as more effective than paid advertising (Iyengar and Kinder 1987), and considerable research has found that media coverage directly affects impressions of candidates and issues (Just, Crigler, Alger, and Cook 1996; Krosnick and Kinder 1990). Of course, perceptions of media bias do matter and are increasing, and viewers condition their responses to media content based upon their assessment of the source (Miller and Krosnick 2000). On the other hand, DellaVigna and Kaplan (2007) tracked the introduction of Fox into a variety of markets after 1996, and found that communities exposed to Fox News demonstrated a 3 to 28 percent increase in Republican vote share when compared to communities that were not yet exposed to the network. Research also finds that individual level vote decisions can be influenced by the endorsements of candidates by political groups or newspaper editorials (Lau and Redlawsk 2001; Erikson 1976). Kahn and Kenney (2002), for instance, find that candidates endorsed by a newspaper's editorial page tend to receive comparatively better coverage on the news side of the operation, highlighting the fact that the exact mechanism by which endorsements are effective remains unclear. Although most of the media coverage research has focused on news programs, presidential candidates have been increasingly likely to serve as guests on entertainment programs in recent years, campaign behavior that may be impacting voters. In a study of the 2000 election, Baum (2005) found politically inattentive viewers of entertainment oriented talk shows reported more favorable evaluations of opposing party candidates, and were more likely to vote for those candidates than non viewers or more politically engaged viewers (but see Prior 2003). Morris (2009) found viewers of the Daily Show, a daily comedy show which bases many of its jokes on observation of the political process, were differentially affected by the program during the 2004 presidential conventions. The show's content was more derogatory toward Republicans, according to Morris's analysis, and viewers demonstrated lower favorability toward George W. Bush but not John Kerry after being exposed to the program. Given the increasing fragmentation of the media environment, future research needs to give greater attention to the way evolving media options have changed candidate strategy, campaign exposure, and campaign effects. Directions for Future Research Although the aforementioned research has offered clear and compelling evidence that campaigns matter, there are important questions still to be addressed in the field. (p. 214) Looking back over the last decade of research, it is particularly striking how issues of data quality are intimately intertwined with the study of campaign effects. The dominance of the American National Election Study (ANES), with its pre and post election design, made it difficult to capture campaign dynamics or to look at changes in opinions from early in the campaign. With the emergence of new academic surveys, we have seen an expansion in the type of research questions that can be answered. New panel surveys have allowed the identification and measurement of individual level change in candidate preference (for example, Hillygus and Jackman 2003). The rolling cross sectional studies just a few hundred people interviewed every day have helped to isolate movements in response to particular campaign events (for example, Johnston, Hagen, and Jamieson 2004). Just as important has been the recognition that we must take seriously issues of measurement. Self reported campaign exposure, in particular, has been the subject of considerable scrutiny both because highly engaged and sophisticated respondents are more likely to seek out campaign exposure and because campaigns are more likely to target those same individuals (Gerber and Green 2000a). Recent research has leveraged creative research designs to get a handle on these concerns. Some have turned to experiments to avoid the problems of observational research. The benefits of experiments are well known. Experiments allow scholars to generate causal inferences through carefully constructed comparisons of groups receiving different treatments. The key question is whether the conditions that created the effects in the experiment will ever be replicated in the real world (Morton and Williams 2008). Even the innovative and largescale field experiments must contend with the concern that the treatments are not comparable to what voters experience from an actual campaign. Beyond this well known concern, Gaines, Kuklinski, and Quirk (2007) caution that certain experimental designs can potentially contaminate inferences. When there are multiple experiments Page 7 of 12

conducted within a single survey, the first experimental treatment respondents encounter can have spillover effects during administration of the second experiment. They also point out that the discovery of treatment effects may be valuable at the theoretical level, but significant experimental effects do not necessarily lead to substantive consequences for political behavior. Other innovative research has looked for exogenous sources of variation in campaign treatments so called quasi or natural experiments. For example, Huber and Arceneaux (2007) use the arbitrary shapes of television markets and the incidental exposure of voters in non battleground states to presidential advertisements in order to test whether and how these ads shape voter behavior. Beyond issues of measurement, scholars should be sensitive to the fact that other data errors can undermine campaign effects research. It is well recognized, for instance, that surveys are facing increasing methodological challenges declining response rates, increasing numbers of cell phone only households, and so on. Burden (2000) finds that turnout overreporting in the ANES has gotten worse in recent decades because of declining response rates among individuals least likely to participate. In response, some scholars have turned to nonprobability samples: non probability Internet surveys (p. 215) make it possible to obtain a large number of cases for very little money. But a convenience sample, in which the probability that an individual is selected into the sample cannot be computed, will not allow for generalizations no matter its size (Couper and Miller 2008; AAPOR 2010). Even experimental designs can be compromised by the use of a non probability sample if, for instance, the effect of the experimental treatment is conditional on respondent characteristics such as political sophistication, interest, or knowledge. While methodological issues will always be central to the study of campaign effects, the field should also be careful not to devolve into a purely methodological debate that pushes theoretical development into the background. Indeed, the field needs more theory driven studies and a stronger structural framework for developing hypotheses. What are the mechanisms by which campaign information can shape the vote decision? What are the links between attitudes, knowledge, and behavior? How do context and institutions shape the campaign effects that are possible? More generally, the field could benefit from interdisciplinary research on persuasion and attitude change, which shows that persuasion depends on the properties and interaction of message, source, receiver, and channel factors. All of these factors determine how someone responds to information encountered in the campaign and will shape any campaign effects that can be observed. Finally, the field must remain cognizant of the fact that our empirical investigations speak to many of the same normative controversies motivating the early scholars of campaign effects. For instance, we know that campaigns generally promote political learning, but we know less about whether this learning serves to reduce or expand the knowledge gap between citizens from different social strata. We know that campaign messages can induce voters to give more weight to some predispositions over others in their decision making, but we do not know how that changes the dynamics of the relationship between voters and candidates and whether it is a benefit or detriment to representation. And we know that campaign efforts can increase the likelihood that some voters go to the polls, but we do not know whether that ultimately improves the health of democracy. Answering these important normative questions will require moving beyond the simplistic debates about whether or not campaigns have any influence on the electorate to more careful consideration of the ways in which voters learn and incorporate campaign information and the consequences for the dynamics of the relationship between candidates and citizens in representative democracy. Bibliography References AAPOR (American Association of Public Opinion Research). 2010. AAPOR Report on Online Panels, Mar. At http://www.aapor.org/am/template.cfm?section=aapor_committee_ and_task_force_reports&template=/cm/contentdisplay.cfm&contentid=2223. Accessed June 3, 2010. (p. 216) ALTHAUS, S., NARDULLI, P., and SHAW, D. 2002. Candidate Appearances in Presidential Elections, 1972 2000. Page 8 of 12

Political Communication, 19/1: 49 72. ANSOLABEHERE, S., and IYENGAR, S. 1995. Going Negative: How Political Advertisements Shrink and Polarize the Electorate. New York: Free Press. ARCENEAUX, K., and NICKERSON, D. 2009. Who Is Mobilized to Vote? A Re Analysis of Seven Randomized Field Experiments. American Journal of Political Science, 53/1: 1 16. BARTELS, L. 1993. Messages Received: The Political Impact of Media Exposure. American Political Science Review, 87: 267 85. 2006. Priming and Persuasion in Presidential Campaigns. In Capturing Campaign Effects, ed. H. E. Brady and R. Johnston. Ann Arbor: University of Michigan Press. BAUM, M. 2005. Talking the Vote: Why Presidential Candidates Hit the Talk Show Circuit. American Journal of Political Science, 44: 213 34. BECK, P., DALTON, R. J., GREENE, S., and HUCKFELDT, R. 2002. The Social Calculus of Voting: Interpersonal, Media, and Organizational Influences on Presidential Voting. American Political Science Review, 96/1: 57 73. BERELSON, B., LAZARSFELD, P., and MCPHEE, W. 1954. Voting. Chicago: University of Chicago Press. BIMBER, B., and DAVIS, R. 2002. Campaigning Online: The Internet in U.S. Elections. New York: Oxford University Press. BRADER, T. 2005. Striking a Responsive Chord: How Political Ads Motivate and Persuade Voters by Appealing to Emotions. American Journal of Political Science, 49/2: 388 405. BRIANS, C. L., and WATTENBERG, M. 1996. Campaign Issue Knowledge and Salience: Comparing Reception from TV Commercials, TV News, and Newspapers. American Journal of Political Science, 40: 172 93. BURDEN, B. 2000. Voter Turnout and the National Election Studies. Political Analysis, 8: 389 98. CAMPBELL, A., CONVERSE, P., MILLER, W., and STOKES, D. 1960. The American Voter. New York: Wiley. CAMPBELL, J. 2000. The American Campaign: U.S. Presidential Campaigns and the National Vote. College Station: Texas A&M University Press. CHERRY, L., and WINK, K. 1992. The Convention Bump. American Politics Research, 20: 287 307. CHAIKEN, S., LIBERMAN, A., and EAGLY, A. H. 1989. Heuristic and Systemic Information Processing within and beyond the Persuasion Context. In Unintended Thought, ed. J. S. Uleman and J. A. Bargh. New York: Guilford Press. CLINTON, J., and LAPINSKI, J. 2004. An Experimental Study of Political Advertising Effects in the 2000 Presidential Election. Journal of Politics, 66/1: 67 96. CONVERSE, P. 1964. The Nature of Belief Systems in Mass Publics. In Ideology and Discontent, ed. D. E. Apter. New York: Free Press. COUPER, M., and MILLER, P. 2008. Web Survey Methods. Public Opinion Quarterly, 72/5: 831 5. DELLAVIGNA, S., and KAPLAN, E. 2007. The Fox News Effect: Media Bias and Voting. Quarterly Journal of Economics, 122/3: 1187 234. DRUCKMAN, J. N., and NELSON, K. R. 2003. Framing and Deliberation: How Citizens' Conversations Limit Elite Influence. American Journal of Political Science, 47/4: 729 45. ERIKSON, R. 1976. The Influence of Newspaper Endorsements in Presidential Elections. American Journal of Political Science, 20: 207 34. FINKEL, S. E. 1993. Reexamining the Minimal Effects Model in Recent Presidential Elections. Journal of Politics, 55: Page 9 of 12

1 21. (p. 217) FRANZ, M., FREEDMAN, P. B., GOLDSTEIN, K. M., and RIDOUT, T. N. 2007. Campaign Advertising and American Democracy. Philadelphia: Temple University Press. FREEDMAN, P., FRANZ, M., and GOLDSTEIN, K. 2004. Campaign Advertising and Democratic Citizenship. American Journal of Political Science, 48: 723 41. GAINES, B. J., KUKLINSKI, J. H., and QUIRK, P. J. 2007. The Logic of the Survey Experiment Reexamined. Political Analysis, 15/1: 1 20. GEER, J. G., and LAU, R. R. 2006. A New Approach for Studying Campaign Effects. British Journal of Political Science, 35: 269 90. GELMAN, A., and KING, G. 1993. Why Are American Presidential Election Campaign Polls So Variable When Votes Are So Predictable? British Journal of Political Science, 23: 409 51. GERBER, A., and GREEN, D. 2000a. The Effects of Canvassing, Telephone Calls, and Direct Mail on Voter Turnout: A Field Experiment. American Political Science Review, 94: 653 63. 2000b. The Effects of a Nonpartisan Get Out the Vote Drive: An Experimental Study of Leafletting. Journal of Politics, 62/3: 846 57. GREEN, D., and GERBER, A. 2004. Get Out the Vote: How to Increase Voter Turnout. Washington, DC: Brookings Institution. and NICKERSON, D. 2003. Getting Out the Vote in Local Elections: Results from Six Door to Door Canvassing Experiments. Journal of Politics, 65: 1083 96. HILLYGUS, D. S., and HENDERSON, M. 2010. Policy Issues and the Dynamics of Vote Choice in the 2008 Presidential Election. Journal of Elections, Public Opinion and Parties, 20/2: 241 69. and JACKMAN, S. 2003. Voter Decision Making in Election 2000: Campaign Effects, Partisan Activation, and the Clinton Legacy. American Journal of Political Science, 47: 583 96. and SHIELDS, T. 2008. The Persuadable Voter: Wedge Issues in Presidential Campaigns. Princeton: Princeton University Press. HOLBROOK, T. 1996. Do Campaigns Matter? Thousand Oaks, CA: Sage. 1999. Political Learning from Presidential Debates. Political Behavior, 21/1: 67 89. 2002. Presidential Campaigns and the Knowledge Gap. Political Communication, 19: 437 54. and MCCLURG, S. 2005. The Mobilization of Core Supporters: Campaigns, Turnout, and Electoral Composition in United States Presidential Elections. American Journal of Political Science, 49/4: 689 703. HUBER, G., and ARCENEAUX, K. 2007. Identifying the Persuasive Effects of Presidential Advertising. American Journal of Political Science, 51/4: 957 77. HUCKFELDT, R., and SPRAGUE, J. 1995. Citizens, Politics, and Social Communication: Information and Influence in an Election Campaign. New York: Cambridge University Press. IYENGAR, S., and KINDER, D. 1987. News that Matters: Television and American Opinion. Chicago: University of Chicago Press. and SIMON, A. 2000. New Perspectives and Evidence on Political Communication and Campaign Effects. Annual Review of Psychology, 51: 149 69. JOHNSTON, R., BLAIS, A., BRADY, H. E., and CRETE, J. 1992. Letting the People Decide: Dynamics of a Canadian Election. Stanford, CA: Stanford University Press. Page 10 of 12

HAGEN, M. G., and JAMIESON, K. H. 2004. The 2000 Presidential Election and the Foundations of Party Politics. Cambridge: Cambridge University Press. JUST, M., CRIGLER, A., ALGER, D., and COOK, T. 1996. Crosstalk: Citizens, Candidates, and the Media in a Presidential Campaign. Chicago: University of Chicago Press. (p. 218) KAHN, K., and KENNEY, P. 1999. The Spectacle of U.S. Senate Campaigns. Princeton: Princeton University Press. 2002. The Slant of the News: How Editorial Endorsements Influence Campaign Coverage and Citizens' Views of Candidates. American Political Science Review, 96: 381 94. KEY, V. O., JR. 1966. The Responsible Electorate. Cambridge, MA: Harvard University Press. KRAMER, G. 1970. The Effects of Precinct Level Canvassing on Voter Behavior. Public Opinion Quarterly, 34: 560 72. KRASNO, J. S., and GREEN, D. 2008. Do Televised Presidential Ads Increase Voter Turnout? Evidence from a Natural Experiment. Journal of Politics, 70/1: 245 61. KROSNICK, J. A., and KINDER, D. R. 1990. Altering the Foundations of Support for the President through Priming. American Political Science Review, 84: 497 512. LANOUE, D. J. 1995. The 1992 Presidential Debates in Focus. Public Opinion Quarterly, 59: 448 50. and SCHROTT, P. 1991. The Joint Press Conference. New York: Greenwood. LASSWELL, H. 1927. Propaganda Technique in World War I. Boston: MIT Press, 1971. LAU, R. R., and REDLAWSK, D. P. 2001. Advantages and Disadvantages of Cognitive Heuristics in Political Decision Making. American Journal of Political Science, 45: 951 71. and ROVNER, I. B. 2009. Negative Campaigning. Annual Review of Political Science, 12: 285 306. SIGELMAN, L., and ROVNER, I. B. 2007. The Effects of Negative Political Campaigns: A Meta Analytic Reassessment. Journal of Politics, 69/4: 1176 209. LAZARSFELD, P., BERELSON, B., and GAUDET, H. 1948. The People's Choice. New York: Sloan and Pearce. LINN, S., MOODY, J., and ASPER, S. 2009. Explaining the Horse Race of 2008. PS: Political Science and Politics, 42/3: 459 65. MASKET, S. 2009. Did Obama's Ground Game Matter? The Influence of Local Field Offices during the 2008 Presidential Election. Public Opinion Quarterly, 73/5: 1023 39. MENDELBERG, T. 2001. The Race Card: Campaign Strategy, Implicit Messages, and the Norm of Equality. Princeton: Princeton University Press. MILLER, J. M., and KROSNICK, J. 2000. News Media Impact on the Ingredients of Presidential Evaluations: Politically Knowledgeable Citizens Are Guided by a Trusted Source. American Journal of Political Science, 44: 301 15. MONSON, J. Q. 2004. Get On Television vs. Get On the Van: GOTV and the Ground War in 2002. In The Last Hurrah? Soft Money and Issue Advocacy in the 2002 Congressional Elections, ed. D. B. Magleby and J. Q. Monson. Washington, DC: Brookings Institution. MORRIS, J. 2009. The Daily Show with Jon Stewart and Audience Attitude Change during the 2004 Party Conventions. Political Behavior, 31/1: 79 102. MORTON, R., and WILLIAMS, K. 2008. Experimentation in Political Science. In The Oxford Handbook of Political Methodology, ed. J. M. Box Steffensmeier, H. E. Brady, and D. Collier. New York: Oxford University Press. Page 11 of 12

MUTZ, D. 2006. Hearing the Other Side: Deliberative versus Participatory Democracy. New York: Cambridge University Press. NICKERSON, D. 2007. Does Email Boost Turnout? Quarterly Journal of Political Science, 2: 369 79. NIE, N., VERBA, S., and PETROCIK, J. 1976. The Changing American Voter. Cambridge, MA: Harvard University Press. PRIOR, M. 2003. Any Good News in Soft News? Political Communication, 20: 149 72. (p. 219) 2007. Post Broadcast Democracy: How Media Choice Increases Inequality in Political Involvement and Polarizes Elections. Cambridge: Cambridge University Press. ROSENSTONE, S. 1983. Forecasting Presidential Elections. New Haven: Yale University Press. and HANSEN, J. M. 1993. Mobilization, Participation, and Democracy in America. New York: Longman. SCHROTT, P. R. 1990. Electoral Consequences of Winning Televised Campaign Debates. Public Opinion Quarterly, 54: 567 86. SHAW, D. 1999. The Effect of TV Ads and Candidate Appearances on Statewide Presidential Votes, 1988 1996. American Political Science Review, 93: 345 61. 2006. The Race to 270: The Electoral College and the Campaign Strategies of 2000 and 2004. Chicago: University of Chicago Press. VERBA, S., SCHLOZMAN, K. L., and BRADY, H. 1995. Voice and Equality: Civic Voluntarism in American Politics. Cambridge, MA: Harvard University Press. Notes: (1) Although television advertising receives the bulk of the attention from candidates and scholars, there are a few studies of radio advertising (for example, Panagopoulos and Green 2008). Bradford H. Bishop Bradford H. Bishop is a graduate student in political science at Duke University. D. Sunshine Hillygus D. Sunshine Hillygus (Ph.D., Stanford University) is an Associate Professor of political science and Director of the Initiative on Survey Methodology at Duke University. Her research and teaching specialties include public opinion, political behavior, survey research, campaigns and elections, and information technology and society. She is co-author of The Hard Count: The Social and Political Challenges of the 2000 Census (Russell Sage Foundation, 2006) and The Persuadable Voter: Wedge Issues in Political Campaigns (Princeton University Press, 2008). Page 12 of 12