An Investigation of Political Identity Congruence in the Real Estate Industry

Introduction

Similar to many countries worldwide, the United States currently finds itself trapped in an era of extreme political polarization. Parties appear to be in constant conflict due to numerous deeply-rooted and highly-committed contradictory perspectives (Cox & Kernell, 2019). The divergence of intellectual and political leaders has grown to the point where there is now little ideological overlap between opposing parties (Bartels, 2016; Desilver, 2022). Similarly, polarization within the general public is also heightened (Gentzkow, 2016), and many Americans now actively dislike members of the opposing political party more than they like their own co-partisans (Iyengar & Krupenkin, 2018; Parker & Janoff-Bulman, 2013). Polarization not only affects attitudes toward fellow citizens, but it also affects consumption behaviors. Partially based on the influence of business communications and social media discussions, American consumers prefer purchasing brands whose political positions are aligned with their own (Duman & Ozgen, 2018; Kelm & Dohle, 2018; Matos, Vinuales, & Sheinin, 2017). This preference may extend from brands to services, and to the people who provide those services (Mercurio & Aiken, 2022). How you vote appears to have a direct relationship to how you shop, where you shop, and which products you purchase (Ordabayeva & Fernandes, 2018).

In real estate, service quality is primarily a function of tangible organizational factors and agent characteristics (Seiler, Webb, & Whipple, 2000). Organizational factors include the non-personal, ancillary components of service delivery. Agent characteristics deal with personal character traits. Traditionally, agent characteristics have been denominated by personal traits, such as reliability and responsiveness, that contribute to the alignment of customer expectations and service deliverables (Johnson, Dotsm, & Dunlap, 1988; Nelson & Nelson, 1995; Parasuraman, Zeithaml, & Berry, 1985, 1988; Seiler et al., 2000). However, given today’s political climate and the growth of political consumerism in which consumers seek ideological accord with purchased brands (Dalakas, Melancon, & Szczytynski, 2023), real estate clients might also use political ideology to assess agents.

Political consumerism involves using product or service purchases to signal personal adherence to a given political ideology or social stance (Cambefort & Pecot, 2020; Jung & Mittal, 2020; Matos et al., 2017). We contend that consumers seek out political identity congruence (PIC) through their purchase and consumption behaviors. That is, consumers wish to align themselves in terms of political values thereby strengthening their desired identities (Reed et al., 2012). This behavior is predicted by balance theory which suggests that individuals will seek congruence between personal ideals and the public, a sociopolitical signaling of corporate brands or service providers (Dalakas & Levin, 2005; Heider, 1958; Lee, Motion, & Conroy, 2009). To enable this balanced self-congruence, real estate clients may seek agents whose political positions are aligned with their own. Hence, the emergence of PIC.

The purpose of this work is to examine whether (and to what extent) PIC takes place in the residential real estate industry. In particular, does PIC inform clients’ selection and assessment of real estate agents? If so, how should real estate agents respond? Understanding whether PIC exists and the way in which it affects the choice and assessment of real estate agents will help individual agents better determine if or how they should publicly communicate their social values and political positions. If PIC exists in real estate sales interactions, an agent might vary communication strategies to facilitate that agent’s appeal.

Literature Review

Political consumers are individuals who are more politically active than the average and who are often interested in lifestyle-related social issues (Baek, 2010). They have a higher propensity to broadcast their political positions and to use their purchases to influence social change (Ward & de Vreese, 2011; Webster, 1975). Political consumers also generally prefer to purchase products or services from brands whose expressed ideologies mirror their own (Matos et al., 2017; Sandikci & Ekici, 2009). This notion of ideological congruence between salesperson and client is the essence of PIC. These preferences extend to the locations and methods associated with purchase and consumption behaviors (Duman & Ozgen, 2018; Ordabayeva & Fernandes, 2018).

The collective actions of political consumers are referred to as political consumerism, which can be defined as the use of purchase or consumption behaviors to engage in civic or political activism (see Dalakas et al., 2023; Farah & Newman, 2010; Maniates, 2001; Stolle, Hooghe, & Micheletti, 2005). Though originally addressing the intersection between consumer behavior and environmental activism, the concept of political consumerism now extends beyond the environment to include a broad array of political and sociocultural ideals (Dalakas et al., 2023; Farah & Newman, 2010; Maniates, 2001). Political consumerism is a powerful force since it is closely related to personal identity, both as a result of identity formation and as a means by which to strengthen identity (see Micheletti & Stolle, 2007). Additionally, its strength is furthered by today’s politically polarized landscape (Tsai, Yuan, & Coman, 2023; Zorell, 2018) and the ease of message dissemination fostered by corporate digital communications and social media (Kelm & Dohle, 2018). Specifically, consumers who are exposed to and interact with organizational communications regarding political constructs are more likely to network with like-minded individuals and engage in boycotting or buycotting behaviors (see Cheng, Zhang, & Gil de Zuniga, 2023). Boycotting involves withholding business from organizations that broadcast incongruent values, while buycotting involves rewarding organizations that communicate congruent positions (Neilson, 2010). These behaviors are moderated by the perceived personal relevance of the communicated value or issue (You, Hon, & Lee, 2023).

The need for PIC between parties in a real estate sales situation has yet to be thoroughly researched. Few real estate firms have experienced consumer boycotts. However, the presence of political consumerism, which is intimately linked to PIC (Carney, Jost, Gosling, & Potter, 2008; DellaPosta, Shi, & Macy, 2015; Gromet, Kunreuther, & Larrick, 2013; Kaikati, Torelli, Winterich, & Rodas, 2017), is expected in the real estate market. First, residential real estate implies dealing with peoples’ homes, which have been found to be an important extension of consumers’ personal identities (Belk, 1988; Proshansky, Fabian, & Kaminoff, 2014; Somerville, 1997). When personal identities are involved, political consumerism is likely present (Micheletti & Stolle, 2007). Second, real estate agents serve as clients’ representatives (Day & Nourse, 1991). In essence, agents become an extension of a client’s identity. Since agents represent clients, agent/client identities may become merged, and the importance of client/agent congruence may be amplified. Political consumerism attaches great importance to ideological congruence (Duman & Ozgen, 2018), and this importance is augmented in real estate where clients must seek ideological accord not only with an organizational brand, but also with the individual agent who personally represents the client.

Additionally, relatively few studies have examined the impact of politics on real estate, and no studies have specifically examined the way in which the need for PIC impacts the choice and assessment of real estate agents. Traditionally, the choice and assessment of real estate agents has been examined via the SERVQUAL and RESERV models (Nelson & Nelson, 1995; Parasuraman et al., 1985, 1988; Seiler et al., 2000). The seminal SERVQUAL model denominated service quality as a function of tangible organizational features and personal agent characteristics such as reliability, responsiveness, assurance, and empathy. Even though Johnson et al. (1988) determined that real estate agents can be accurately assessed via SERVQUAL’s agent characteristic dimensions, Nelson and Nelson (1995) created the Real Estate Service Quality (RESERV) scale to specifically adapt SERVQUAL for the real estate industry. Neither SERVQUAL nor RESERV specifically consider PIC, but they do provide a foundation for the current study through confirming that consumers use specific personal attributes to choose and assess real estate agents. For example, agent knowledge is an important choice criterion, especially when consumers are new to an area or unfamiliar with the home buying process (Gibler & Nelson, 2003; Kaynak, 1985). Likewise, an agent’s ability to understand a client’s needs and desires is important (Kidwell, Hardesty, Murtha, & Sheng, 2011). Additionally, an agent’s emotional intelligence is related to customers’ assessments of that agent (Hoxha & Zeqiraj, 2022). According to Goleman (1995), emotional intelligence includes four specific competencies (self-awareness, social awareness, self-management, and relationship management), and each of these competencies influences realtor choice and assessment (Danquah & Wireko, 2014; Halim, 2021; Kidwell et al., 2011; Landy, 2005; Oluwatofunmi & Amietsenwu, 2019; Swanson & Zobisch, 2014). Therefore, we hypothesize:

Apart from specific real estate agent personality characteristics, additional variables impact the choice and assessment of real estate professionals. When given the opportunity, consumers generally prefer to use within-network agents (DiMaggio & Louch, 1988). Within-network agents are agents who enjoy pre-existing social ties with consumers, and these agents are more likely to be chosen and to receive positive service assessments from customers (DiMaggio & Louch, 1988). Generally, race and gender are not considered when choosing or assessing agents (see Benjamin, Chinloy, Jud, & Winkler, 2007). However, it is unclear whether political positions impact agent choice and assessment since no studies have specifically examined whether PIC affects selection and assessment. Nevertheless, since personal identity is a key ingredient in both political consumerism and home selection, a connection may exist. Both homes and agents can be viewed as extensions of a consumer’s personal identity. Therefore, the desire for self-congruence, which is at the heart of political consumerism, may be expected when choosing and assessing real estate agents. Thus, we predict:

Finally, the notion of political partisanship likely plays a critically important role in the relationship between salesperson and client. Partisanship is defined as a set of beliefs and feelings that culminate in a sense of psychological attachment to a political party (Huddy, Mason, & Aaroe, 2015). Researchers in the field of political science use an expressive approach that assesses an internalized sense of party membership to measure partisanship (Oscarsson & Holmberg, 2020). We believe that highly partisan consumers will likely be more inclined to accept and reward (as well as avoid and punish) realtors that they perceive as congruent (incongruent) with their strong party affiliations. Hence, hypothesize:

Methods

This study utilized a 2 × 2 between-subjects experimental design (gender of realtor by political party affiliation of realtor). Data collection took place over the Internet using the online recruiting service Prolific. With such services, there is always the potential for misrepresentation across a non-representative sample (Sheehan, 2018). Moreover, the contrived nature of the experiment is not reflective of a realistic sales encounter; however, the research design seemed most appropriate as it allows for the control and isolation of key study variables. Further, the benefits of Internet samples include reduced costs, quick responses, increased participant diversity, superior data quality, and greater research flexibility (Hulland & Miller, 2018). In this instance, we sought to measure the opinions of a nationwide sample of adults. Subjects were screened for being U.S. citizens over 30 years old, with the assumption that age would bring forth more established political viewpoints as well as a greater likelihood of home shopping/buying experiences.

The survey had four major sections. First, respondents were introduced to the research, told of the survey progression, and asked to imagine they were actively searching for a mid-sized home to purchase. They were asked to provide consent, and then they were exposed to a professionally designed home rendering. Further, to increase realism, the rendering included a sales-oriented paragraph that generically described the home as a “blend of comfort and convenience”, built with a “tradition of excellence”, sized “1,840 square feet with 3 bedrooms and 2 baths”, and “resting in a friendly neighborhood”. The paragraph increased the realism of the online sales-environment and also served to stimulate involvement.

Section two of the survey began with two attention-check questions regarding the home’s door color and exterior. These two questions would later serve to expose respondents who had not properly filled out the survey. Respondents then read one of four treatments that described the potential realtor. All four paragraphs described the realtor as dedicated, with 20 years of experience and a member of the local Chamber of Commerce. Treatments varied on pronouns (“she” or “he”) and political affiliation (“Democratic” or “Republican”). Respondents were then asked two dichotomous questions about political consumerism and their preferences.

Section three of the survey asked subjects to rate 15 personal trait adjectives assigned to the experimental realtor. On a 1–7 scale of “Not at all descriptive” to “Very descriptive” subjects rated adjectives drawn from: Aaker, 1997; Duman and Ozgen, 2018; Goldberg, 1990; 1995; Mercurio and Aiken, 2022; and, Nelson and Nelson. All 15 traits were positively valenced attributes of effective salespeople.

Finally, the survey’s fourth section asked questions about respondents’ political party membership, levels of partisanship, and various demographics. Huddy et al.’s four-item political partisanship scale (2015) utilized 7-point ratings and helped distinguish levels of political identification. The scale yielded a Cronbach’s alpha value of 0.91, indicating an excellent level of reliability (George & Mallery, 2003). Accordingly, a new partisanship variable was computed as the mean of the four scale items. Categorical demographic data included gender, age groups, education, and income levels.

Results

The initial sample included 640 subjects; however, 56 were eliminated due to inordinately short response times and/or errors made on attention-check questions (n = 584). Regarding party affiliation, the sample distribution was 20.5% Republican (n = 120), 44.7% Democrat (n = 261), and 34.1% Independent/Other (n = 199) (with four subjects not identifying). Additionally, the sample was 47.8% (n = 279) female, 58.7% (n = 343) were in their 30s, and a full 56.3% (n = 329) had college degrees. Income categories were somewhat spread out with no group falling below 10% and the highest number of subjects (40.9%; n = 239) falling in the $50,000–$100,000 range.

Tests revealed that while political consumerism and PIC do exist in real estate sales, sexism and gender preferences generally do not. In answering the reward-based question, “I am more likely to use a realtor that holds the same political views that I do” the data revealed a significant relationship to political party (χ² (2) = 32.4; p < .01). More Democrats answered affirmatively and more Republicans answered negatively (compared to expectations). With regards to the punishment-based question, “I prefer not to do business with realtors who make political statements I disagree with” the test again showed significance (χ² (2) = 36.7; p < .01). Democrat respondents reported that they were more likely to engage in politically-based punishment. However, Republicans and Independents were more likely (than expected) to answer this question negatively. Thus, H1 was supported (see Tables 1, 2). When asked about choosing a realtor based on the salesperson’s gender, tests by party, age, education, and income were not significant. Although, in testing subjects’ preferred realtor gender by subject gender, the data revealed that women have a significant preference for female realtors (χ² (2) = 13.2; p < .01). Yet, it should be noted that a large majority of subjects reported that they had “no preference” regarding their realtor’s gender (supporting Benjamin et al., 2007).

Assessing trait-adjective perceptions across the entire sample, analysis of variance (ANOVA) was run to compare the means of each of the four experimental conditions (see Table 3). Taking the entire sample in aggregate, the group rated the experimental realtors as significantly different in 13 of the 15 adjectives (F-statistics ranged from 2.4 to 152.6; p < .06). While the Republican realtors were viewed as more traditional and conservative, the Democrat realtors (regardless of assigned gender) rated higher on practically all other adjectives (with the one exception of relatively equal ratings of “business savvy”). Further, the woman Republican realtor condition scored the lowest on several positively-valenced adjectives (specifically: ethical, trustworthy, exciting, intelligent, competent, sophisticated, friendly, organized, a good communicator, and fiscally responsible).

Noting that the above tests were somewhat skewed by including the large number of Independent subjects, the goal of the next phase of analysis was to isolate key differences between Republican and Democrat subjects. Thus, the next tests removed Independents from analyses. Evaluating only Republican subjects, four significant differences were uncovered. The Republican realtor conditions were rated as more friendly, traditional, conservative, and business savvy (t-values ranging from 2.0 to 9.5; p < .05). Then, controlling for only Democrat subjects, all 15 trait adjectives showed significant differences between the two Republican versus the Democrat realtor conditions. Here Democrat subjects rated the Democrat realtors as significantly higher on 13 adjectives and significantly lower in being traditional and conservative (t-values ranging from 3.9 to 14.8; p < .01). Accordingly, H2 was largely supported. See Table 3.

Next, the research team sought to evaluate the influence of partisanship on realtor perceptions and political consumerism. Several tests revealed that higher levels of partisanship corresponded to greater levels of political consumerism as well as greater differences in perceptions (of the experimental realtors). Overall, the more highly partisan the subject the greater the tendency towards political consumerism (“Rewarding” χ² (2) = 17.3 and “Punishing” 48.2; p < .01). Additionally, more highly partisan subjects showed greater differences in perceptions of trait adjectives (with the exceptions of sophisticated, traditional, and conservative). Lastly, all correlation coefficients were positive (between ratings and partisanship scores) and t-values ranged from 2.0 to 3.5; p < .05 (after grouping respondents into high and low levels of partisanship). Hence, H3a and H3b were supported.

Finally, various demographic analyses revealed a few key relationships amongst the variables. In terms of gender analyses, women were more likely to desire female realtors. Further, Democratic women were more likely to both reward (same-party realtors) and punish (opposite-party realtors). This finding is especially pronounced amongst more highly-partisan subjects. Income analyses showed that all incomes groups were equally likely to exhibit preferences for PIC. Yet, across the four income groups, a significant inverted U-shape pattern emerged wherein the lowest income group (annual income under $50,000) rated the positively-valenced adjectives lowest, then the means ascended as income groups ascended, and then then highest income group rated the trait adjectives low again (F-values ranged from 3.1 to 9.1; p < .05). See Table 4. This inverted U-shape appeared in tests of 12 of the 15 adjectives. In essence, the lowest and highest income groups seemed to express the most negative views towards realtors. Lastly, across the four education groups (from high school, some college, bachelor degree, graduate degree), the same inverted U pattern was seen in 11 of the 15 adjectives. The middle two subject groups rated the experimental realtors significantly higher (F-values ranged from 3.2 to 6.0; p < .05).

Discussion

The purpose of this study was to determine if consumers desire PIC in the residential real estate industry. Specifically, does political congruence between agents and customers influence a client’s selection and assessment of an agent? Results indicate that PIC does exist and that it impacts both the selection and assessment of agents.

Overall, perceptual variance between Republicans and Democrats has been extensively documented (see Bartels, 2002; Bullock, Gerber, Hill, & Huber, 2013; Gerber & Huber, 2010, Jerit & Barabas, 2012). The observed variance, however, has been most closely associated with assessments of economic or political ideas. Until recently, it has been unclear whether perceptual variance of these impersonal constructs primarily stems from divergent factual assessments of situational factors or is also influenced by conscious political cheerleading (Bartels, 2002; Brennan & Lomasky, 1997; Hamlin & Jennings, 2011; Jerit & Barabas, 2012; Schuessler, 2000; Shapiro & Bloch-Elkon, 2008). Recent research suggests both constructs bear influence (Bullock et al., 2013; Iyengar, Lelkes, Levendusky, Malhotra, & Westwood, 2019). The current study’s findings add nuance to the contemporary understanding of perceptual differences and also move the discussion into the personal sphere.

First, the current study confirms that perceptual differences are not exclusively driven by divergent factual assessments. Democrat respondents positively assessed the expected performance of politically congruent agents before the performance occurred. This introduces the notion of bias and its effects on agent selection and performance analysis. Could this bias be a form of political cheerleading? Possibly. However, the authors suggest that further research is required to unearth all causal factors behind this observed bias.

Second, the current study moves the discussion of bias into the personal realm. Most prior research has reviewed political or ideological bias from an impersonal perspective. For example, political bias has been observed in the assessment of impersonal political or economic matters such as crime rates, unemployment, inflation, the federal deficit, and public sector performance (Bartels, 2002; Conover, Feldman, & Knight, 1986; Marvel, 2016). Recently, researchers have begun exploring whether these impersonal biases have personal counterparts. For example, Nicholson, Coe, Emory, and Song (2016) confirmed that assessments of another person’s physical attractiveness are influenced by political positions. The current study continues this personal exploration and confirms significant gender effects (found by Mercurio & Aiken, 2022; Nicholson et al., 2016). In both studies, females exhibited stronger personal bias based on political congruence than males. The current study found female Democrats were most likely to use PIC to select or avoid real estate agents. This result supports prior research indicating that females are generally more idealistic than males (Donoho, Heinze, & Kondo, 2012; Karande, Rao, & Singhapakdi, 2002; Schminke & Ambrose, 1997; Wang & Calvano, 2015).

In relation to the real estate industry, the current study provides insights concerning both agent selection and agent evaluation. Regarding agent selection, respondents were more likely to choose an agent who shared similar political positions, and less likely to choose an agent who held dissimilar views. Unsurprisingly, this effect strengthened as levels of partisanship increased. The more partisan a respondent, the more likely they were to select a politically similar agent. The effect was particularly pronounced among Democrat women who were most likely to engage in selection reward or punishment based on an agent’s political beliefs. The effect was less evident among Democrat men and was missing among Republicans. These results support the contention that Democrats are more ideologically demanding of PIC than Republicans (who are more practical and results-oriented) (Kirkpatrick, 1975). Similarly, the current study found that Democrat clients were more likely to select agents according to ideological accord. Conversely, Republicans were more pragmatic, seeking to find the right house at the right price regardless of politics.

Practical implications for real estate agents are clear. Republican agents should not communicatively signal their political leanings, especially concerning positions for which consumers assign high personal relevance (see You et al., 2023). This is particularly true for female Republican agents. Signaling their partisanship will not necessarily attract fellow Republican clients, and it will reduce the likelihood of attracting Democrat clients. On the other hand, Democrat agents will potentially benefit from broadcasting their political positions. Republican clients will generally ignore this signaling and will instead focus on performance characteristics, while Democrat clients will use the signaling as a basis for agent selection.

Apart from selecting agents, the assessment of agents is likewise affected by political positions. Out of 15 agent characteristics, Republican respondents assessed Republican real estate agents more favorably on four characteristics (friendly, traditional, conservative, and business savvy). In contrast, Democrat respondents assessed Democrat agents more favorably on 13 of 15 characteristics. Practical implications can be drawn from the fact that female Democrats disproportionally utilized political positions to influence realtor performance assessments. This finding uniquely illustrates prior gender research regarding the use of perceptual cues in high involvement purchase situations. In high involvement purchase situations, females disproportionally rely on peripheral cues to make assessments while men rely more heavily on central cues (Stros, Heinze, Möslein-Tröppner, & Lim, 2020). Since a realtor’s political position has no rational bearing on his/her ability to negotiate a successful real estate deal, utilizing political position to influence one’s assessment of that realtor is tantamount to assessing a product based on the peripheral cues of an associated advertising campaign. Therefore, real estate agents should consider their political signaling to be a peripheral cue. Since these cues are less likely to influence men and more likely to influence women in high involvement purchase situations (Stros et al., 2020), agents should carefully consider political broadcasts. If their political positions lean left, they can freely signal their positions so long as they also include central cues to attract males from both political parties. However, if they lean right, political signaling should be kept to a minimum since it will not attract Republicans and it will repel female Democrats.

A final implication is the need for greater recognition of the impact of bias across non-related constructs. While political positions are logically connected to many personal practices and assessments (see Hersh & Goldberg, 2016 for a discussion of physicians’ political positions and patient care practices), the current study highlights an arena where the link between political positions and demonstrable performance is not readily apparent. Basing selection and assessment of real estate agents on PIC offers a troubling picture of the nature of implicit and explicit bias.

There are several limitations and areas for future research. First, although the current study provides a needed examination of the personal impact of political consumerism, its findings are not generalizable across all interpersonal situations. The findings accord with several recent studies indicating that political positions influence assessments of non-related personal factors such as attractiveness (Nicholson et al., 2016), but more research is required to understand the relative extent of this effect. Second, it should be noted that the experimental nature of this work made for an unnatural and contrived sales situation. The research design allowed for subjects to concentrate on variables and it allowed the research team to analyze variables in isolation; however, it was admittedly unrealistic. Third, an area for future research would lies in the larger scope of political power. That is, data were collected during a time when Democrats largely controlled the U.S. government. It would be interesting to conduct future research when the Republican party was in control.

Understanding gender effects is another area for future research. The current study confirmed gender effects but did not deeply explore them. In particular, a model of PIC is needed that explores the impact of gender and political partisanship on selection and assessment bias.

Conclusion

Amid rising polarization, PIC has become increasingly visible. Companies have responded by taking public stances on political and social issues. In real estate, individual agents may wonder if or how they should follow suit. The current study examined PIC in the residential real estate market and determined that consumers use political positions to both select and assess real estate agents. However, since the effect is primarily seen among Democrat customers, Republican real estate agents are unlikely to realize returns by signaling their political ideologies.