The Effects of Candidate’s Ethnic Source Cue and Party Affiliation on Hispanic Voters’ Homophily Perceptions and Voting Intent

Abstract

Because of the rapid demographic shifts in the United States (US), it is vital to examine how voters of different ethnic backgrounds perceive the ethnic source cue of a candidate in political advertising and how such perceptions influence voting behaviour. We used a post-test-only experiment to investigate the effects of candidate’s ethnic source cue and party affiliation on voters’ homophily perceptions and voting intent. A total of 335 college students from a large southwestern university in the US were recruited to take part in a post-test-only experiment. Results from multivariate analysis of covariance (MANCOVA) found that candidate’s ethnic source cue and party affiliation have significant main effects on voters’ homophily perceptions and voting intent. The moderating effects of voters’ own ethnicity, party identification and gender were also found to have statistically significant effects on voters’ homophily perceptions and voting intent. Theoretical and practical implications are discussed.

Keywords

Political advertising ethnic source cue party affiliation homophily perceptions party identification voting intent experimental design

The Emergence of Hispanic Constituency in the US

The consensus among national news media and public opinion research institutes was that minority voters will play a significant role in the 2012 US election (Fox and Frey 2012). Among many ethnic minorities, Hispanic voters have attracted much attention to better understand their voting behaviours. Among 10 per cent Hispanic electorate in the 2012 election, only 23 per cent of Hispanic constituents voted for the Republican Party (LatinoVoteMap.org, 2012). A pre-election survey conducted by Gallup predicted that the ethnic minorities turnout and enthusiasm was likely to determine if President Obama will be re-elected in 2012 (cited in Fox). Because the Democratic Party’s agenda was often viewed to favour more comprehensive federal subsidy for housing, education and medical care, President Obama was perceived as more advantageous than his opponent, former Massachusetts Governor Romney, among the Hispanic voters (Frey, 2012). The sentiment was widespread, as observed by one Republic Party’s campaign volunteer when talking to the Hispanic voters in Denver, Colorado, who received the feedback, ‘Mitt Romney’s for the millionaires. We’re these poor Hispanics, so we’re going to vote for Obama because he’s for the little guy’ (Abdullah, 2012). The political behaviour of ethnic minorities has thus increasingly drawn the attention of political advertising researchers and practitioners.

It was not until the late 1990s that scholars (for example, Alvarez & Bedolla, 2003; Subervi-Vélez & Connaughton, 1999) began to examine the political behaviour of Hispanics. However, there is a growing consensus among researchers and practitioners that this important demographic segment cannot simply be approached as part of the dominant Anglo-American culture and categorized as one homogenous group due to their immigration patterns, culture, assimilation and language use (Alvarez & Bedolla, 2003; Holland & Gentry, 1999; Subervi-Vélez, 1986; Torres & Gelb, 2002).

The US Census Bureau has reported that more than one-third of the population belongs to minority groups and almost one-sixth of its residents are Hispanics (approximately 46.9 millions) (as cited in Christie, 2009). Census data have shown that Hispanic population has reached 16.3 per cent of the US population of 308 million (as cited in Humes, Jones & Ramirez, 2011). Hispanic share of voters has seen substantial growth among several key southwestern states in two presidential elections (2004 and 2008), according to Pew Research analysis (as cited in Lopez, 2008). In New Mexico, the share of Hispanic voters has grown from 32 per cent in 2004 to 41 per cent in 2008. The percentages of Hispanic voters in Arizona, Nevada and Colorado have also increased 4–5 per cent from 2004 to 2008 (as cited in Lopez, 2008). Considering the mounting political power of Hispanics, the Democratic Party and then Senator Barack Obama launched a 20 million outreach campaign to mobilize Hispanic voters in their 2008 presidential campaign (Davis, 2008). The campaign produced anticipated results. CNN’s exit polls in Arizona, California, Colorado, Florida, Illinois, Nevada, New Jersey, New Mexico and Texas found that 67 per cent of the Hispanics voted for Democrat Barack Obama, while 31 per cent of them voted for Republican John McCain (as cited in Lopez, 2008).

Since the 1960s, studies have shown that every Democratic presidential candidate has received the majority of the Hispanic vote. However, the Democratic Party cannot take the Hispanic vote for granted because not all Hispanic subcultures vote in the same manner (Alvarez & Bedolla, 2003; Subervi-Vélez, 2008). Alvarez and Bedolla (2003) found that throughout the years, most Hispanic subcultures have reported being Democratic (67 per cent of Mexican-origin respondents, 66 per cent of Puerto Ricans and 57 per cent of Central American origin) and approximately 69 per cent of Cuban-origin voters reported being Republican. In the 2004 presidential election, various pre-election surveys and exit polls showed that regardless of age, sex, education and income level, the majority of Hispanics (approximately 60 per cent) said they voted for Kerry (Greenberg & Hogan, 2004; Leal, Barreto, Lee & de la Garza, 2005). Historical data over the past four presidential elections (1992–2004) have concluded that Hispanics often voted for Democratic presidential candidates (Hurst, 2012; Lopez & Minushkin, 2008).

In response of the emergence of Hispanic political power, an important question arises: Will a candidate’s ethnic source cue play any role in explaining the voting behaviour among Hispanic voters?

Literature Review

Research has found that political advertising leads to elevated voter turnout (Cappella, Jamieson, Romer & Nurick, 2000; Romer, Jamieson & Cappella, 2000). For this reason, politicians are willing to spend millions of dollars on advertising; in the 2000 election, George Bush and the Republican Party spent over $65 million in television advertising, while Al Gore and the Democrats spent over $61 million (Marks, 2000). Total advertising in 2004 reached $1.85 billion by both political parties (Hau, 2007). The 2012 presidential election alone was estimated to be $2.23 billion, according to Bloomberg (as cited in Hickey, 2012).

Our study is situated with political advertising research and intends to examine whether a candidate’s ethnic source cue and party affiliation influence Hispanic voting behaviour. Specifically, the experimental study examines if and how ethnic source cue and party affiliation of a hypothetical candidate in a political advertisement influence Hispanic participants’ homophily perceptions and voting intent.

Ethnic Source Cue in Political Advertising

Maldonado and Muehling (2006) broadly define cues as stimuli in an advertisement that may develop a deeper association or meaning to the person viewing the advertisement. Researchers have often categorized cues into verbal (for example, words, accent and language use) and non-verbal forms (for example, posture and hand gestures) (Bailenson, Garland, Iyengar & Yee, 2006; Glascock & Ruggiero, 2004; Grier & Brumbaugh, 1999; Holland & Gentry, 1999; Len-Ríos, 2002; Torres & Gelb, 2002). Advertising research has expanded the definition of non-verbal cues to include group-level visual cues such as race and ethnicity (for example, Simpson, Snuggs, Christiansen & Simples, 2000; Valentino, Hutchings & White, 2002).

In this study, ethnic source cue of a candidate is treated as an independent variable. Past researchers, such as Appiah (2001a, 2001b), Brumbaugh (2002) and Grier and Brumbaugh (1999), have found that the physical traits (for example, skin colour, gender, facial structure and stature) of the source (that is, actors, spokespersons, models, etc.) of an advertisement and the accompanying elements contained in the advertisement (that is, non-source elements such as symbols, material objects and dress) can be used as cues. Research has also shown that the ethnicity of the source in an advertisement has an impact on subjects’ attitude towards the advertisement and their purchase intent of a particular product (for example, Green, 1999; Simpson et al., 2000; Whittler, 1989; Whittler & Spira, 2002). Deriving from Bandura’s (2001) social cognitive theory, researchers have attempted to explain the effects of ethnicity of a communication source, arguing that people pay more attention to a source with the same ethnic background and identify more with the source. Green, Simpson et al. and Whittler and Spira found that when ethnic subjects who identified strongly with their own ethnic culture viewed advertisements featuring someone from their own culture, there was a positive effect on the evaluation of the advertisement and the product. Green (1999) and Simpson et al. (2000) also found a significant effect between ethnic source cue and purchase intent of the product. Whittler (1989) found that while black subjects preferred advertisements with black sources than white sources, white participants reacted as favourably to black as to white sources.

Because of the increasing importance of minority groups in the American political landscape, a critical question for political advertising researchers and practitioners is whether a candidate’s own characteristics, along with many message factors, can influence minority voting behaviour. Communication researchers such as Subervi-Vélez (2008) and Subervi-Vélez and Connaughton (1999) have confirmed the importance of mass communication and campaign strategies on Hispanic voting behaviour.

Political advertising researchers have rarely explored how a political candidate’s ethnic source cue will influence voters’ homophily perceptions, as well as the relationship between voters’ homophily perceptions and subsequent voting behaviour. An empirical assessment of these relationships helps understand the voting behaviour of constituencies, no matter whether they belong to minority or mainstream groups. Therefore, to address the paucity in the existing political advertising literature, the purpose of this experimental study is to examine the above-mentioned relationships.

Homophily Perceptions (of a political candidate)

Homophily, a term coined by Lazarsfeld and Merton (1954), is defined as ‘a tendency for friendships to form between those who are alike in some designated respect’ (p. 23). Elias, Appiah and Gong (2011) have defined homophily as ‘the phenomenon of perceived similarity between two people’. From a communication perspective, homophily is perceived as similarity between source and receiver (Allen & Post, 2004; Glascock & Ruggiero, 2006; McCroskey, McCroskey & Richmond, 2006; Simpson et al., 2000). Those with similar lifestyles, beliefs and positions tend to create relationships among each other (Cho, 2005).

Past literature review has established the persuasive influence of source cues on perceived similarity (Allen & Post, 2004). In the political arena, research has shown that similarities between a candidate and voters may help a candidate win the election by relying on personal outreach to Hispanics to encourage them to vote for and support particular candidates (Mcllwain, 2007). Researchers in political advertising have further demonstrated the effects of a candidate’s ethnic source cue on advertising effectiveness and subsequent voting behaviour (Green, 1999; Simpson et al., 2000; Whittler, 1989; Whittler & Spira, 2002). Therefore, the following research question was proposed:

Research Question 1: Will a candidate’s ethnic source cue affect voters’ homophily perceptions?

Hellweg (1979) found that homophily was made up of different dimensions. McCroskey et al. (2006) argued that homophily should be grouped into two primary dimensions: background homophily (for example, perceived similarities as a result of social class, geographic region and ethnicity); and attitude homophily (for example, perceived similarities because of values, behaviour and ideas). Allen and Post (2004) found that, among many of the variables they studied, attitude homophily rather than background homophily was more affected by source cues of a candidate. Allen and Post demonstrated that the perceived expertise and competence of the source affected voters’ attitudes more. Therefore, the following research question was proposed:

Research Question 2: Will a candidate’s ethnic source cue affect different types of voters’ homophily perceptions?

Voting Intent

Empirically oriented voting research dates back to the 1920s, with the advent of survey research: academic surveys in the 1920s and surveys from commercial polling firms in the 1930s (Lewis-Beck, Jacoby, Norpoth & Weisberg, 2008). Historically, voting research has followed two basic paths: if a person will vote (voter turnout); and how a person will vote (voting intent) (Cho, 2005; Lewis-Beck et al., 2008; Yoon, Pinkleton & Ko, 2005). Cho (2005) has operationally defined voting behaviour by asking a voter to indicate candidate choice as simply, ‘Who did you vote for?’ Therefore, this research will focus on how a person will vote. There is a gap in the literature to establish if ethnic source cue will influence voting intention. However, Appiah (2001a, 2001b) have found a close relationship between ethnic source cue and response to advertising. On the basis of the above-mentioned literature, the following research question was proposed:

Research Question 3: Will a candidate’s ethnic source cue affect voters’ voting intent?

Voters’ Party Identification and Demographics as Moderating Variables

The ethnicity-based prediction of voters’ attitudes and voting behaviour is overly simplistic about the relationship between ethnicity and voting behaviour. Therefore, it is logical to speculate that there are other determinants in voters’ decision-making process. For example, Abrajano, Nagler and Alvarez’s (2005) natural experiment in the 2001 City of Los Angeles elections found that a candidate’s position on issues and ideology could also play a more determinant role in voters’ choice. Therefore, except for the homophily that is generated from a match between a candidate’s and voter’s ethnicity, we also consider other variables to explain voters’ behaviour.

A voter’s own identification to a particular party has played a major role and in the voting booth, the most obvious connection of a candidate is that to his or her party (Lewis-Beck et al., 2008). Party identification has been defined as ‘an individual’s affective orientation to an important group–object in his environment’ (Campbell, Converse, Miller & Stokes, 1960, p. 121). Lewis-Beck et al. (2008) stated, ‘party identification is a psychological identification with the party’ (p. 112). In addition, they state that this identification strongly affects voters’ attitudes and voting behaviour. However, past studies have found that voters’ own party identification also play a significant role in the Hispanic vote, with most Hispanics identifying themselves as Democrats (Alvarez & Bedolla, 2003; Cho, 2005).

Research Question 4: Will voters’ own party identification moderate the relationship between a candidate’s ethnic source cue, voters’ homophily perceptions and their voting intent?

In addition to party identification, demographic variables such as gender and ethnicity could also play an important role. Therefore, the fifth research question proposed was:

Research Question 5: Will voters’ demographics (that is, gender and ethnicity) moderate the relationship between a candidate’s ethnic source cue, voters’ homophily perceptions and their voting intent?

Method

The purpose of this research is to establish a causal inference (that is, cause-and-effect relationship) between two independent variables (a candidate’s ethnic source cue and party affiliation) and three dependent variables (attitude homophily perceptions, background homophily perceptions and voting intent) in political advertising. Voters’ own party identification and other two demographic variables (voters’ gender and ethnicity) were also considered to examine their moderating effects on the relationships between independent and dependent variables. Participants’ age was not used in the model because of highly similar age distribution among college students in the sample.

We developed a 2 × 2 post-test-only between subject experiment design to examine the main effects of a candidate’s ethnic source cue and party affiliation on voters’ attitude and background homophily perceptions, as well as their voting intent. The ethnic source cue of a candidate was operationalized at two levels: Hispanic and Black/African American, while a candidate’s party affiliation was also operationalized at two levels: Republican and Democrat. The candidate’s party affiliation was another independent variable and operationalized as Republican and Democrat. The design enabled the researcher to examine the main effects of two independent variables and, at the same time, investigate their interaction effects on participants’ voting intent and two types of homophily perceptions.

Variable Selection and Instrumentation

Ethnic Source Cues

Ethnic source cues have been conceptually defined as changing the skin tone and physical features of the source (Appiah, 2001a, 2001b; Brumbaugh, 2002; Whittler, 1989). This can be done by digitally manipulating the source to match known features of different groups. This can also be done by creating advertisements in which the sources are different people from different ethnic groups, but have the same stature, pose, sex and facial expression (Brumbaugh, 2002; Whittler, 1989). In this study, we used one aspect of ethnic source cue by focusing on a candidate’s ethnicity.

On the basis of past operationalization’s, the researchers developed four advertisements varying on the ethnic source cue of the candidate in the advertisements. This independent variable was manipulated through the use of fictitious print advertisements that were designed specifically for this study. Four full-colour 8½″ × 11″ photographic advertisements of political candidates were developed by varying candidate’s ethnic source cue (Hispanic or Black/African American) and party affiliation (Republican and Democrat). Care was taken in the construction of these advertisements to ensure that the advertisement setting, actors’ physical size, facial expressions and clothing were similar. Thus, the only difference in the advertisement was the candidate’s ethnic source cue. In addition to the photo of the candidate in the test advertisement, the remainder of the advertisement consisted of a patriotic-themed header and footer, with the contents of the advertisements running down the left side of the advertisement. The contents were separated into three headings: business, experience and community. Under each heading were one to three sentences describing the candidate’s background, values and community involvement. The candidate’s party affiliation was stated in the experience section by indicating that the candidate has ‘has always stood up for principled Republican/Democratic values’. The Democratic donkey and Republican elephant were used at the top right-hand side of the advertisements to visually show a candidate’s party affiliation. In summary, the four advertisements were created exactly the same, except for the candidate’s ethnicity and party affiliation, varying in the following combinations: Republican African American/Black candidate, Democratic African American/Black candidate, Republican Hispanic candidate and Democratic Hispanic candidate. These advertisements were created for a hypothetical city councilman election on 2 May (see Figure 1 from the sample advertisement).

Figure 1.

Source: Authors’ own.

Homophily Perceptions

Homophily perceptions have been conceptually defined as ‘perceived similarity between two people’ (Elias et al., 2011). Those that have similar backgrounds, lifestyles and beliefs tend to create bonds with each other (Cho, 2005). Operationally, researchers have employed various techniques to measure homophily (Lazarsfeld & Merton, 1954). However, one of the most used and respected measure’s of homophily is the one first created by McCroskey, Richmond and Daly (1975), and later modified by McCroskey et al. (2006). As mentioned earlier, their studies found that homophily measurements can be grouped into two primary dimensions: background homophily (for example, social class, geographic region and ethnicity) and attitude homophily (for example, values, behaviour and ideas). The perceived homophily scale that was developed by McCroskey et al. (1975) and later modified by McCroskey et al. (2006) was adapted for use in the current research. The modified scale had four items for the background dimension (α = 0.83) (Mean = 2.81, SD = 0.85) and six items for the attitude dimension (α = 0.91) (Mean = 2.65, SD = 0.73) and was measured on a 5-point Likert statement. A composite score for each type of homophily perceptions was computed for later statistical analyses.

Voting Intent

In political advertising literature, voting intent is conceptually defined as the willingness for a subject to vote for a particular candidate (Yoon et al., 2005). In this research, the participants were asked to respond to the following 5-point Likert statement: ‘After reading the ad, I would vote for the candidate’.

Voters’ Own Party Identification

Voters’ own party identification is conceptually defined as the political party one typically identifies with (Campbell et al., 1960). Operationally, voters’ party identification was measured, modifying Valentino et al. (2002), by asking respondents the following question: ‘Generally speaking, do you consider yourself a Democrat, Independent, or Republican?’

Sampling Plan and Characteristics

Non-probability convenience sampling was conducted to allow for the maintenance of a smaller budget and efficiency to execute the experiment (Hocking, Stacks & McDermott, 2003). Furthermore, because our experiment is interested in exploring the causal relationship among the study variables, internal validity was emphasized over the possibility to generalize the results. The true experiment with random assignment was conducted in several undergraduate-level classes consisting of a good mix of lower-division and upper-division students. A total of 335 participants took part in this experiment. About 50.7 per cent (N = 170) of the sample was male, while 49.3 per cent (N = 165) was female. Seventy-five of them (22.4 per cent) said they were Independents; 61.2 per cent (N = 205) classified themselves as Democrats; and 55 (16.4 per cent) said they belonged to Republicans. In terms of their ethnicity, most of them said they were Hispanic/Latino (N = 290, 87.9 per cent) and 12.1 per cent of them (N = 40) indicated that they were not Hispanic or Latino.

Experimental Procedure

Subjects were told that they are participating in an advertising study designed to determine the types of advertisements they like best for a fictitious political campaign for a city councilman on 2 May. Extra credits were given to participants. After briefly explaining the purposes of the study, and procedures, subjects were randomly assigned an instrument that contained one of the four experimental conditions and a questionnaire. Random assignment procedure used in this study enabled the researcher to remove any pre-existing bias among participants and ensure internal validity to better assess the effects of independent variables (Hocking et al., 2004). As this study was only minimally intrusive, debriefing of the respondents was not necessary. At the conclusion of the questionnaire respondents were thanked for their participation.

At the end of the data collection period, 80 (23.9 per cent) of the participants were exposed to the Republican African/Black candidate, while 85 (25.4 per cent) of them saw the Democratic African/Black candidate; and 22.4 per cent (N = 75) of the participants were exposed to the Republican Hispanic candidate, while 28.4 per cent (N = 95) saw the Democratic Hispanic candidate.

Findings

We conducted multivariate analysis of covariance (MANCOVA) to assess the main effects of two independent variables (that is, a candidate’s ethnic source cue and party affiliation) on three dependent variables (that is, participants’ attitude and background homophily perceptions, as well as their voting intent), when taking into consideration the effects of three moderating variables (participants’ own party identification, ethnicity and gender). The following empirical results provided answers to all five research questions.

Statistically significant results concluded that there was a main effect of candidate’s ethnic source cue on three dependent variables (Wilks’ Lambda = 0.89, F = 13.58, p = 0.00 < 0.001) (Research Question 1). The main effect of a candidate’s party affiliation on the dependent variables was also found to be statistically significant (Wilks’ Lambda = 0.86, F = 16.64, p = 0.00 < 0.001). The interaction effect of two independent variables on these dependent variables was also found to be statistically significant in the empirical data (Wilks’ Lambda = 0.91, F = 10.44, p = 0.00 < 0.001). The moderating effects of voters’ own party identification (Research Question 4) and demographics (Research Question 5) on their attitude and background homophily perceptions, as well as their voting intent, were also found to be significant (refer to Table 1).

Table 1.

MANCOVA Results

	Effect	Value	F	d.f.	p-value
Candidate’s Ethnic Source Cue	Pillai’s Trace	0.11	13.58	3.00	0.00***
	Wilks’ Lambda	0.89	13.58	3.00	0.00***
	Hotelling’s Trace	0.13	13.58	3.00	0.00***
	Roy’s Largest Root	0.13	13.58	3.00	0.00***
Candidate’s Party Affiliation	Pillai’s Trace	0.14	16.64	3.00	0.00***
	Wilks’ Lambda	0.86	16.64	3.00	0.00***
	Hotelling’s Trace	0.16	16.64	3.00	0.00***
	Roy’s Largest Root	0.16	16.64	3.00	0.00***
Candidate’s Ethnic Source Cue × Candidate’s Party Affiliation	Pillai’s Trace	0.09	10.44	3.00	0.00***
	Wilks’ Lambda	0.91	10.44	3.00	0.00***
	Hotelling’s Trace	0.10	10.44	3.00	0.00***
	Roy’s Largest Root	0.10	10.44	3.00	0.00***
Candidate’s Ethnic Source Cue × Voters’ Ethnicity	Pillai’s Trace	0.10	5.70	6.00	0.00***
	Wilks’ Lambda	0.90	5.80	6.00	0.00***
	Hotelling’s Trace	0.11	5.90	6.00	0.00***
	Roy’s Largest Root	0.11	11.12	3.00	0.00***
Candidate’s Party Affiliation × Voters’ Party Identification	Pillai’s Trace	0.23	13.84	6.00	0.00***
	Wilks’ Lambda	0.77	14.51	6.00	0.00***
	Hotelling’s Trace	0.29	15.17	6.00	0.00***
	Roy’s Largest Root	0.27	28.39	3.00	0.00***
Candidate’s Ethnic Source Cue × Voters’ Party Identification × Voters’ Gender	Pillai’s Trace	0.29	8.42	12.00	0.00***
	Wilks’ Lambda	0.73	8.64	12.00	0.00***
	Hotelling’s Trace	0.33	8.74	12.00	0.00***
	Roy’s Largest Root	0.21	16.70	4.00	0.00***

Source: Authors’ analysis.

Note: *** p < 0.001.

While the MANCOVA model examined the main effect of a candidate’s ethnic source cue on three dependent variables simultaneously, when taking into consideration the moderating effects of voters’ demographics, analysis of covariance (ANCOVA) test further showed the main effect of a candidate’s ethnic source cue on individual dependent variables to provide results for Research Questions 1–5. A candidate’s ethnic source cue had main effect on voters’ attitude homophily perceptions (F = 4.08, p = 0.04 < 0.05) (Research Questions 1 and 2) and voting intent (F = 19.75, p = 0.00 < 0.001) (Research Question 3), but not their background homophily perceptions (F = 1.04, p = 0.31 > 0.05) (Research Question 2). The candidate’s party affiliation was found to affect only participants’ voting intent (F = 39.42, p = 0.00 < 0.001) (Research Question 4). There were also statistically significant interaction effects of candidate’s ethnic source cue and party affiliation on voters’ background homophily perceptions (F = 11.20, p = 0.00 < 0.001) and voting intent (F = 26.94, p = 0.00 < 0.001), but not their attitude homophily perceptions (F = 1.50, p = 0.22 > 0.05). The moderating effects of voters’ ethnicity, party identification and gender were also found to be statistically significant on their attitude homophily perceptions (F = 7.24, p = 0.00 < 0.001) (Research Questions 4 and 5), background homophily perceptions (F = 2.40, p = 0.05 < 0.05) (Research Questions 4 and 5) and voting intent (F = 14.75, p = 0.00 < 0.001) (Research Questions 4 and 5) (refer to Table 2).

Table 2.

ANCOVA Results

Independent Variables (or moderating variables)	Dependent Variables	Type III Sum of Squares	d.f.	Mean Square	F	p-value
Candidate’s Ethnic Source Cue	Attitude HomophilyBackground HomophilyVoting Intent	1.79	1	1.79	4.08	0.04*
		0.70	1	0.70	1.04	0.31
		12.50	1	12.50	19.75	0.00***
Candidate’s Party Affiliation	Attitude HomophilyBackground HomophilyVoting Intent	0.14	1	0.14	0.31	0.58
		0.09	1	0.09	0.13	0.72
		24.94	1	24.94	39.42	0.00**
Candidate’s Ethnic Source Cue × Candidate’s Party Affiliation	Attitude HomophilyBackground HomophilyVoting Intent	0.66	1	0.66	1.50	0.22
		7.57	1	7.57	11.20	0.00***
		17.04	1	17.04	26.94	0.00***
Candidate’s Ethnic Source Cue × Voters’ Ethnicity	Attitude HomophilyBackground HomophilyVoting Intent	4.38	2	2.19	4.98	0.01**
		4.40	2	2.20	3.25	0.04*
		0.97	2	0.48	0.76	0.47
Candidate’s Party Affiliation × Voters’ Party Identification	Attitude HomophilyBackground HomophilyVoting Intent	9.30	2	4.65	10.58	0.00***
		8.85	2	4.43	6.55	0.00***
		51.75	2	25.87	40.90	0.00***
Candidate’s Ethnic Source Cue × Voters’ Party Identification × Voters’ Gender	Attitude HomophilyBackground HomophilyVoting Intent	12.73	4	3.18	7.24	0.05*
		6.47	4	1.62	2.40	0.00***
		37.33	4	9.33	14.75	0.00***

Source: Authors’ analysis.

Note: * p < 0.05; ** p < 0.01; *** p < 0.001.

Discussion and Conclusion

This study explores whether a candidate’s ethnic source cue and party affiliation affect voters’ homophily perceptions and voting intent in political advertising. Our empirical findings confirmed that a candidate’s ethnic source cue in political advertising plays an important role in influencing voters’ attitude homophily perceptions and voting intent of the candidate—both are likely to predict their actual voting behaviour. When considering voters’ own ethnicity, party identification or gender, candidate’s ethnic source cue continues to be a critical determinant in explaining voters’ homophily perceptions and voting intent. Because the formation of homophily perceptions is contingent on the perceived similarity and congruence between the source and the receiver, our empirical findings have also supported the importance of target marketing practices to ensure the effectiveness of political campaigns. In this study, we empirically examined the effects of these targeting practices when political advertisers attempt to create a sense of congruence and homophily between the message source (that is, a political candidate) and receivers (that is, voters).

The Effect of Ethnic Source Cue on Voters’ Homophily Perceptions and Voting Intent

The first three research questions attempt to empirically examine whether congruence with a candidate’s ethnic source cue will influence voters’ homophily perceptions and voting intent. In other words, that Hispanic subjects would perceive the political candidate to be more similar and be more likely to vote when viewing a political advertisement with a Hispanic candidate as opposed to viewing an advertisement with a black candidate. In our study, we found that candidate’s ethnic source cue affects voters’ attitude homophily perceptions and voting intent of a candidate. Our findings are in line with previous research (Brumbaugh, 2002; Whittler & Spira, 2002) and support the importance of candidate’s ethnic source cue in political advertising effectiveness.

In the advertising literature, the source is one of the most widely noticed elements (Whittler & Spira, 2002). Source cues in political advertising are highly visually salient and are processed quickly and automatically (Brumbaugh, 2002). The salience of these cues has been found to influence the viewers’ processing of the message since it attracts attention to the source (Whittler, 1989). Even though a candidate’s ethnic source cues are often viewed as peripheral cues in the processing of advertising messages (Whittler & Spira, 2002), it is likely that voters are more motivated to pay attention to advertisements featuring a candidate with the same ethnic background as their own. We speculate that when both the source and the receivers have the same ethnic background, voters tend to perceive the candidate to be similar, as reflected in their attitude homophily perceptions and voting intent. Social cognitive theorists have established the importance of source cues in motivating people to attend to a message as a result of identification with the communication source. Given that homophily perception leads to favourable voting intent, the candidate’s ethnic source cue is likely to generate positive voting behaviours among voters, when effectively manipulated in political advertising campaigns.

Much research on source effects on persuasion (Groenendyk & Valentino, 2002) has pointed to similar results. For example, attractive sources have been found to be more persuasive than others at delivering the same message (Chaiken, 1979). Research by Chaiken and Eagly (1983) further found that likeable sources were more persuasive than unlikeable sources. By extending the relationship between candidate’s characteristics to homophily perceptions, this study suggested that Hispanics should react favourably to a Hispanic candidate in a political advertisement. As Subervi-Vélez (2008) points out, using Hispanic spokespersons will be an effective approach for a political party to gain Hispanic support.

The Moderating Effect of Voters’ Party Identification, Ethnicity and Gender

The remaining research questions aim to examine the moderating effects of voter-related variables on the effects of candidate’s ethnic source cue and party affiliation on the relationships among independent and dependent variables. For example, the study has examined if any main effect exists between voters’ own party identification with a candidate’s party affiliation and more favourable homophily perceptions about the candidate, and subsequent voting intent. In other words, a Democrat candidate should generate a higher level of homophily perceptions and voting intent from subjects who identify more with the Democratic Party than with the Republican Party. Results from this research are in accordance with most literature (Connaughton & Jarvis, 2004; Grynaviski, 2006; Lewis-Beck et al., 2008) that suggests that a candidate’s party affiliation does play a role in voters’ voting intent. Since Hispanics have surpassed African Americans as the nation’s largest minority group, it is important for both political parties to understand how party affiliation is related to voting intent among Hispanics (Connaughton & Jarvis, 2004). Our empirical results supported the moderating effects of candidate’s party affiliation and voters’ party identification.

We observed a statistically significant interaction effect between voters’ own ethnicity and a candidate’s ethnic source cue on voters’ perceived attitude and background homophily. As Whittler and Spira (2002) have also found, African American participants tend to engage in biased processing of advertising messages when exposed to an advertisement showing an African American model. We found that when voters perceive a candidate to be the same ethnic background, they will have a high level of homophily perceptions. Even though Valentino et al. (2002) found that candidate’s ethnic source cue does indeed have an impact on voters’ political decision-making process, our results did not find voters’ voting intent is affected by the similarity in the ethnic background between the source and the receivers. It is likely that the mere perceptions of homophily do not necessarily lead to the intention to vote for that candidate; factors such as candidate’s party affiliation, positions in various issues and ideology could play a more critical role.

Theoretical Implications

The results of the present study provided interesting insights regarding the use of ethnic source cue in political advertising as a means of influencing voters’ homophily perceptions and voting intent. Several theoretical implications can be derived from the integration of ethnic source cue with political party affiliation. Elias et al. (2011) used a social identity theory to explain why homophily perception is important in advertising research because ethnic minorities are more motivated to improve their perceptions of self-worth than members of an ethnic majority. Therefore, they will have a strong ethnic identification with the endorser of a product. Social cognitive theorists further help us understand how homophily perceptions contribute to the processing of political advertising. The statistically significant main effects of a candidate’s ethnic source cue on homophily perceptions and voting intent allow us to understand its role in a political advertising context. Moreover, this study also begins to address the issue of whether ethnic source cues are capable of eliciting a response, even when strong political party affiliation is present.

The study also contributed to our understanding and application of voters’ party identification and other demographics. The relationship between voters’ party identification and their voting intent has been found to be inconclusive (Lewis-Beck et al., 2008). Scholars have argued that, at times, it is difficult to determine if a voters’ voting intent is based strictly on party identification or other factors (Lewis-Beck et al., 2008). This research has shown that whenever people are exposed to political advertisements that have been linked with a political party that they align with, they tend to vote for that particular candidate that is representing the party.

Limitations of the Study

The results of this study should be interpreted in light of limitations of the methods used. This study concentrates on a small subset of voters for its research: American undergraduate students conveniently recruited from a large public university. Rozin (2007) calls this very small focus the ‘99% of humanity hole’ since most research focuses on less than 1 per cent of humanity. Using undergraduate students as subjects does have a distinct advantage: they are a convenient sample in which to investigate interactions between variables under highly controlled conditions (Rozin, 2007; Siegel, Alvaro, Crano, Lac, Ting and Jones, 2008). Additional studies using non-college student samples are needed to address these limitations. However, with these caveats in mind, the findings have important theoretical implications.

As in most experimental research on advertising effects, the conditions surrounding exposure to the stimuli were not completely naturalistic. Viewing advertisements contained in an experiment packet and responding to a questionnaire does not necessarily simulate naturalistic exposure conditions (Torres & Briggs, 2007). In addition, the study used only one cue (that is, ethnic source cue), which may raise external validity issues (Torres & Briggs, 2007). This cue is somewhat simplistic. In an effort to isolate the effects of ethnicity, we might have oversimplified the experimental advertising stimuli by using very similar advertisements to avoid confounding effects of different backgrounds, model positioning, dress and so forth. Therefore, future research should examine the effects of ethnic source cues using additional cues (for example, gender, age and lifestyle). Researchers should also explore the effects of language used, media placement and other campaign characteristics that could further support the value of measuring the impact of ethnic source cues.

References

Abdullah

(2012, November 10). With their big political win, the new American electorate has arrived. CNN website. Retrieved 1 April 2013, from http://www.cnn.com/2012/11/09/politics/demographic-political-power/index.html

Abrajano

M.A.

, Nagler

, & Alvarez

R.M.

(2005). A natural experiment of race-based and issue voting: The 2001 City of Los Angeles election. Political Research Quarterly, 58(2), 203–218.

Allen

J.L.

, & Post

D.J.

(2004). Source valence in assessing candidate image in a local election. Communication Research Reports, 21(2), 174–187.

Alvarez

R.M.

, & Bedolla

L.G.

(2003). The foundations of Latino voter partisanship: Evidence from the 2000 election. The Journal of Politics, 65(1), 31–49.

Appiah

(2001a). Black, white, Hispanic, and Asian American adolescents’ responses to culturally embedded ads. The Howard Journal of Communications, 12(1), 29–48.

Appiah

(2001b). Ethnic identification on adolescents’ evaluations of advertisements. Journal of Advertising Research, 41(5), 7–22.

Bailenson

J.N.

, Garland

, Iyengar

, & Yee

(2006). Transformed facial similarity as a political cue: A preliminary investigation. Political Psychology, 27(3), 373–386.

Bandura

(2001). Social cognitive theory of mass communication. Media Psychology, 3(3), 265–299.

Brumbaugh

A.M.

(2002). Source and nonsource cues in advertising and their effects on the activation of cultural and subcultural knowledge on the route to persuasion. Journal of Consumer Research, 29(2), 258–269.

10.

Campbell

, Converse

P.E.

, Miller

M.E.

, & Stokes

D.E.

(1960). The American voter. New York: John Wiley & Sons.

11.

Cappella

, Jamieson

K.H.

, Romer

, & Nurick

(2000). Does political advertising affect turnout? If so, how, when, and for whom? In Jamieson

K.H.

(ed.), Everything you think you know about politics…and why you’re wrong (pp. 93–96). New York: Basic Books.

12.

Chaiken

(1979). Communicator physical attractiveness and persuasion. Journal of Personality and Social Psychology, 39(5), 752–766.

13.

Chaiken

, & Eagly

A.H.

(1983). Communicator modality as a determinant of persuasion: The role of communicator salience. Journal of Personality and Social Psychology, 45(2), 241–256.

14.

Cho

(2005). Media, interpersonal discussion, and electoral choice. Communication Research, 32(3), 295–322.

15.

Christie

(2009, May 14). Hispanic population boom fuels rising U.S. diversity. Retrieved 10 April 2014, from http://articles.cnn.com/2009-05-14/us/money.census.diversity_1_hispanic-population-latinos-and-other-minority-pew-hispanic-center?_s=PM:US

16.

Connaughton

S.L.

, & Jarvis

S.E.

(2004). Apolitical politics: GOP efforts to foster identification from Latinos, 1984–2000. Communication Studies, 55(3), 464–480.

17.

Davis

(2008, July 30). Democrats launch Latino outreach. Wall Street Journal, 252(25), A5–A5.

18.

Elias

, Appiah

, & Gong

(2011). Effects of strength of ethnic identity and product presenter race on black consumer attitudes: A multiple-group model approach. Journal of Interactive Advertising, 11(2), Spring. Retrieved 4 April 2014, from http://jiad.org/article144

19.

Fox

(2012, October 26). Gallup predicts voter turnout will remain unchanged in 2012 election. U.S. News & World Report. Retrieved 3 April 2014, from http://www.usnews.com/news/articles/2012/10/26/gallup-predicts-voter-turnout-will-remain-unchanged-in-2012-election

20.

Frey

W.H.

(2012, May). Why minorities will decide the 2012 U.S. election. Washington, DC: The Brookings Institution. Retrieved 1 April 2014, from http://www.brookings.edu/research/opinions/2012/05/01-race-elections-frey

21.

Glascock

, & Ruggiero

T.E.

(2004). Representations of class and gender on primetime Spanish-language television in the United States. Communication Quarterly, 52(4), 390–402.

22.

Glascock

, & Ruggiero

T.E.

(2006). The relationship of ethnicity and sex to professor credibility at a culturally diverse university. Communication Education, 55(2), 197–207.

23.

Green

C.L.

(1999). Ethnic evaluations of advertising: Interaction effects of strength of ethnic identification, media placement, and degree of racial composition. Journal of Advertising, 28(1), 49–64.

24.

Greenberg

, & Hogan

(2004). Hispanics and the 2004 presidential elections. Democracy Corps. Retrieved 20 April 2014, from http://www.democracycorps.com/reports/analyses/Hispanics_and_the_2004_election.pdf

25.

Grier

S.A.

, & Brumbaugh

A.M.

(1999). Noticing cultural differences: Ad meanings created by target and non-target markets. Journal of Advertising, 28(1), 79–93.

26.

Groenendyk

E.W.

, & Valentino

N.A.

(2002). Of dark clouds and silver linings: Effects of exposure to issue versus candidate advertising on persuasion, information retention, and issue salience. Communication Research, 29(3), 295–319.

27.

Grynaviski

J.D.

(2006). A Bayesian learning model with applications to party identification. Journal of Theoretical Politics, 18(3), 323–346.

28.

Hau

(2007, December 7). Political ad spending set to climb sharply. Forbes.com. Retrieved 29 April 2014, from http://www.forbes.com/2007/12/07/politics-campaign-spending-biz-media-cx_lh_1207bizpolitics.html

29.

Hellweg

S.A.

(1979). An examination of voter conceptualization of the ideal political candidate. Southern Speech Communication Journal, 44(4), 373–85.

30.

Hickey

(2012, December 13). The total cost of the 2012 presidential race was astounding. Business Insider. Retrieved 3 April 2014, from http://www.businessinsider.com/election-cost-2-billion-2012-12

31.

Hocking

J.E.

, Stacks

D.W.

, & McDermott

S.T.

(2003). Communication Research (3rd ed.). Boston, MA: Pearson Educations, Inc.

32.

Holland

, & Gentry

J.W.

(1999). Ethnic consumer reaction to targeted marketing: A theory of intercultural accommodation. Journal of Advertising, 28(1), 65–77.

33.

Humes

K.R.

, Jones

Nicholas A.

, & Ramirez

R.R.

(2011, March). Overview of race and Hispanic origins: 2010. Retrieved 1 April 2014, from http://www.census.gov/prod/cen2010/briefs/c2010br-02.pdf

34.

Hurst

S.R.

(2012, May 3). Obama enjoys overwhelming Hispanic support in 2012 race. Retrieved 3 April 2014, from http://www.thegrio.com/politics/obama-enjoys-overwhelming-hispanic-support.php

35.

LatinoVoteMap.Org (2012, December 10). Latino influence on 2012 election. Retrieved 1 March 2014, from http://www.latinovotemap.org/map/

36.

Lazarsfeld

, & Merton

R.K.

(1954). Friendship as a social process: A substantive and methodological analysis. In Berger

, Abel

& Page

C.H.

(Eds), Freedom and control in modern society (pp. 18–66). New York: Van Nostrand.

37.

Leal

D.L.

, Barreto

M.A.

, Lee

, & de la Garza

R.O.

(2005). The Latino vote in the 2004 election. PS: Political Science and Politics, 38(01), 41–49.

38.

Len–Ríos

M.E.

(2002). The Bush and Gore Presidential Campaign web sites: Identifying with Hispanic voters during the 2000 Iowa caucuses and New Hampshire primary. J&MC Quarterly, 79(4), 887–904.

39.

Lewis-Beck

M.S.

, Jacoby

W.G.

, Norpoth

, & Weisberg

H.F.

(2008). The American voter revisited. Ann Arbor: The University of Michigan Press.

40.

Lopez

M.H.

(2008, November 5). How Hispanic voted in the 2008 election. Retrieved 4 March 4 2014, from http://pewresearch.org/pubs/1024/exit-poll-analysis-hispanics

41.

Lopez

M.H.

, & Minushkin

(2008, July 24). 2008 national survey of Latinos: Hispanic voter attitudes. Washington, DC: Pew Hispanic Center.

42.

Maldonado

, & Muehling

D.D.

(2006). The effect of social identity cues under incidental ad exposure conditions. Journal of Promotion Management, 12(2), 77–91.

43.

Marks

(2000, October 31). The 2000 campaign: The spending on ads; Bush has advertised more in California than in Swing States. The New York Times, Retrieved 11 August 2015 from, http://www.nytimes.com/2000/10/31/us/2000-campaign-spending-ads-bush-has-advertised-more-california-than-swing-states.html

44.

McCroskey

J.C.

, Richmond

V.P.

, & Daly

J.A.

(1975). The development of a measure of perceived homophily in interpersonal communication. Human Communication Research, 1(4), 323–332.

45.

McCroskey

L.L.

, McCroskey

J.C.

, & Richmond

V.P.

(2006). Analysis and improvement of the measurement of interpersonal attraction and homophily. Communication Quarterly, 54(1), 1–31.

46.

Mcllwain

C.D.

(2007, May). Racial identity, ideology, and the youth vote: Observations from the 2004 presidential campaign. American Behavioral Scientist, 50(9), 1231–1238.

47.

Romer

, Jamieson

J.H.

, & Cappella

(2000). Does attack advertising affect turnout? In Jamieson

K.H.

(ed.), Everything you think you know about politics…and why you’re wrong (pp. 107–110). New York: Basic Books.

48.

Rozin

(2007). Exploring the landscape of modern academic psychology: Finding and filling the holes. American Psychologist, 62(8), 754–766.

49.

Siegel

J.T.

, Alvaro

E.M.

, Crano

W.D.

, Lac

, Ting

, & Jones

S.P.

(2008). A quasi-experimental investigation of message appeal variations on organ donor registration rates. Health Psychology, 27(2), 170–178.

50.

Simpson

E.M.

, Snuggs

, Christiansen

, & Simples

K.E.

(2000). Race, homophily, and purchase intentions and the black consumer. Psychology & Marketing, 17(10), 877–889.

51.

Subervi-Vélez

F. A.

(1986). The mass media and ethnic assimilation and pluralism: A review and research proposal with special focus on Hispanics. Communication Research, 13(1), 71–96.

52.

Subervi-Vélez

F. A.

(Ed.). (2008). The mass media and Latino politics: Studies of media content, campaign strategies and survey research: 1984–2004. New York: Routledge.

53.

Subervi-Vélez

F.A.

, & Connaughton

S.L.

(1999). Targeting the Latino vote: The Democratic party’s 1996 mass-communication strategy. In de la Garza

R.O.

& DeSipio

(eds), Awash in the mainstream: Latino politics in the 1996 elections (pp. 47–71). Boulder: Westview Press.

54.

Torres

I.M.

, & Briggs

(2007). Identification effects on advertising response: The moderating role of involvement. Journal of Advertising, 36(3), 97–108.

55.

Torres

I.M.

& Gelb

B.D.

(2002). Hispanic–targeted advertising: More sales? Journal of Advertising Research, 42(6), 69–75.

56.

Valentino

N.A.

, Hutchings

V.L.

, & White

I.K.

(2002). Cues that matter: How political ads prime racial attitudes during campaigns. The American Political Science Review, 96(1), 75–90.

57.

Whittler

T.E.

(1989). Viewers’ processing of actor’s race and message claims in advertising stimuli. Psychology & Marketing, 6(4), 287–309.

58.

Whittler

T.E.

, & Spira

J.S.

(2002). Model’s race: A peripheral cue in advertising messages? Journal of Consumer Psychology, 12(4), 291–301.

59.

Yoon

, Pinkleton

B.E.

, & Ko

(2005). Effects of negative political advertising on voting intention: An exploration of the roles of involvement and source credibility in the development of voter cynicism. Journal of Marketing Communications, 11(2), 95–112.