Further to the right: Uncertainty, political polarization and the American “Tea Party” movement

Abstract

The Tea Party entered U.S. politics in a time of economic uncertainty, positioning itself far to the right of the conservative movement. Its highly conservative position has allowed it to provide a clear self-definition that contrasts with more moderate and liberal political views. To examine the Tea Party's influence on American political prototypes, we manipulated the comparative context in which participants received an extreme pro-normative message from a Tea Party group. Conservatives (N = 47), primed with self-uncertainty, supported the extreme position, indicating more conservative views for both themselves and similar others when primed with an intergroup versus an intragroup context. Results are discussed in terms of the ability for extreme ingroup factions to polarize prototypes under self-conceptual uncertainty.

Keywords::

• Uncertainty
• Polarization
• Minority influence
• Extremism
• Self-categorization

A powerful coalition of Tea-Party-backed freshmen walks the halls of Congress, where they have unsettled the establishment and pushed their own party to the right. On the campaign trail, the support of Tea Party groups has helped launch campaigns and is a force candidates ignore at their peril. (Rachel Streitfeld, CNN Political Producer, 2011)

The “Tea Party” burst on the American political scene during the profound economic and social uncertainty provoked by the 2008 stock market plunge and ensuing global recession. Although a numerical minority, the Tea Party has attracted disproportionate media coverage due to the political impact of its promotion of a fiscally and socially conservative agenda that seemingly places it far to the right of establishment conservative and Republican ideology (see Postel, 2012). Initially this group started with protests held across the USA, its members angrily shouting out against the Obama administration's fiscal policies and taxes, (Trost & Rosenthal, 2012). While its members have ranged from citizens and politicians concerned about taxes and government spending, to those demanding to see the birth certificate of the U.S. president and to see a greater merging of Christianity and politics, its overarching message has remained consistent: the “Taxed Enough Already Party” has focused on the U.S. economy. This movement was successfully able to gain momentum and even win seats in Congress where members formed the “Tea Party Coalition” in an effort to transform the federal government from the inside. Given the prevailing atmosphere of economic uncertainty, will a group such as the Tea Party fragment conservative ideology and attitude steer American politics further to the right, generate a centrist correction, rebalance politics to the left, or have no impact on American politics and political attitudes? In this article we draw on three related literatures (group polarization, minority influence, and uncertainty-identity theory) to derive and test the hypothesis that conservatives experiencing feelings of self-uncertainty shift their attitudes further to the right when exposed to Tea Party rhetoric explicitly aimed at liberals.

Most polls (e.g., Zernicke & Thee-Brenan, 2010) indicate that supporters of the various Tea Party groups (e.g., the Tea Party Express, Tea Party Patriots) tend to be demographically and ideologically homogenous (e.g., Newport, 2010), with members largely identifying themselves as “strongly conservative” (see also Abramowitz, 2012). The purpose of the current research is not a characterization of Tea Party membership; rather, this work focuses strictly on perceptions of the Tea Party (regardless of the actual make-up of the various groups dispersed at both the national and local levels).

We focus this research specifically on the Tea Party's impact on American conservatives, because this allows us to study precisely the group process of interest in the very real climate of uncertainty brought on by tumultuous world and national economies. Specifically, we are interested in the impact of a small group or faction on their larger ingroup. Such factions, which champion ideals that clash with the mainstream of their ingroup, can fracture a group's structure leading to schism (see Sani, 2008; Sani & Pugliese, 2008) if the majority of the overarching group does not embrace their agenda. On the other hand, such a position can lead to conversion if the larger group does embrace their agenda. The latter would leave the larger group intact, but shift its ideology, worldview, and practices further from the center and closer to the extreme faction's position.

According to social identity theory, group identification involves defining and evaluating oneself in terms of the prototypical attributes of one's ingroup, where prototypes capture shared ingroup attributes and accentuate ways in which the ingroup differs from relevant outgroups (Turner, Hogg, Oakes, Reicher, &Wetherell, 1987; see Abrams & Hogg, 2010). Predicated on this idea, the social identity analysis of group-based influence shows that people's behavior conforms to such prototypes (e.g., Abrams & Hogg, 1990; Hogg & Turner, 1987) and thus ingroup prototypes are polarized away from relevant outgroups, often in the direction of more extreme ingroup positions (e.g., Abrams, Wetherell, Cochrane, Hogg, & Turner, 1990; Hogg, Turner, & Davidson, 1990).

In the case of the Tea Party, the polarization prediction is that when conservatives define themselves in contradistinction to the Tea Party, they will polarize away from the Tea Party's views, and will thus express more moderate views. When conservatives do not focus on the Tea Party, but only focus on liberals, they will shift to the right and thus be more aligned with the more extremist Tea Party. However, there is a potential complication here—the Tea Party is an active minority. Research on minority influence shows that active minorities can produce change if the minority is consistent (Mugny, 1982) and is an ingroup rather than an outgroup minority (Crano & Seyranian, 2009). This suggests that the Tea Party will ultimately shift conservatives' attitudes toward the right, as the Tea Party comprises an ingroup minority for conservatives.

The seemingly conflicting psychologies of polarization and of minority influence have been integrated by David and Turner (1999). They conducted a series of studies examining a minority faction's ability to exert influence on a politically charged issue. Specifically, moderate feminists exposed to a message from a radical pro-feminist group became more pro-feminist in an intergroup context (contrast to anti-feminists) than in an intragroup context (moderate feminists). In the intragroup context, moderates moved away from the minority position to maintain a prototypical group position, while in the intergroup context, participants moved toward the ingroup minority because it maximized the differences between the ingroup and outgroup positions.

These findings point to a polarization-related social identity dynamic underlying the influence of extreme ingroup minorities. An ingroup minority can influence their group to the extent that they have a distinctive social identity with which the larger group can identify. In an intergroup comparative context an extreme ingroup minority provides a polarized ingroup prototype that accentuates intergroup distinctiveness. Thus, such a minority effectively influences the rest of the group to embrace its position and become more extreme.

Distinctive and extreme identities may be particularly attractive in times of uncertainty. According to uncertainty-identity theory (Hogg, 2000, 2007, 2012), people find feelings of uncertainty, particularly about themselves, their identity and their own lives, aversive (cf. Van den Bos, 2009), and are thus highly motivated to reduce such uncertainty. People can effectively reduce self-uncertainty by identifying with self-inclusive groups, particularly groups that are subjectively prominent (Grant & Hogg, 2012) and often hold distinctive and extreme positions (Hogg, Meehan, & Farquharson, 2010). Extreme groups provide a prototype that is unambiguous and clearly defined, because such groups strongly enforce ingroup norms that make obvious the boundaries of the group while prescribing their members' attitudes and behaviors. Such groups tend to be high in entitativity (Campbell, 1958; Hamilton & Sherman, 1996) and are well positioned to reduce the uncertainty of their followers (e.g., Hogg, Sherman, Dierselhuis, Maitner, & Moffitt, 2007).

By placing uncertainty-identity dynamics center-stage, the study reported here goes beyond previous polarization work, such as that by David and Turner (1999) described above. We investigate how self-uncertainty affects the impact of an extreme political faction (the Tea Party) attempting to influence the views of conservatives. We examine this as a function of the social comparative context of the extremist minority's message (intragroup versus intergroup). The Tea Party is one small segment of the larger conservative movement, thus our specific analyses examine how conservatives respond to a highly conservative ingroup minority.

While the Tea Party comprises several national groups and many iconic faces (e.g., Glenn Beck, Lou Dobbs) that might not actually represent an extremist movement given the Tea Party's close ties with establishment Republicans, this research focuses on people's perceptions of the Tea Party as a segment of the conservative movement. Self-identified conservatives read a speech given by a leader of the Tea Party. They were told that the speech had been given to a group of Republicans (intragroup context, as Republicans are a mainstream conservative group) or to a group of Democrats (intergroup context as Democrats are a mainstream liberal group). Just before indicating their evaluation of and support for the Tea Party message, the participants were cognitively primed to feel more or less uncertain about themselves (self-uncertainty).

We predicted that conservatives primed to feel self-uncertain and led to believe that the speech that was targeted at Democrats would be most supportive of the message and would likewise be most likely to polarize toward the attitudes espoused by the extreme ingroup position. Moreover, we predicted a third-party effect (e.g., Duck, Hogg, & Terry, 1995; Duck, Terry, & Hogg, 1998; Gardikiotis, 2008) such that conservatives would view the Tea Party message as important for their ingroup only under uncertainty in the intergroup context. This finding would lend further support to the hypothesis that the combination of social comparative context and uncertainty provides an opening for social influence from an extreme wing of the ingroup. These predictions allow us to analyze conservatives' attitudes from multiple perspectives (i.e., participants' self-reports of their own attitudes as well as their reports of how the message impacts a relevant group membership).

Method

Participants

Students at an urban public university in Texas provided data at two time points as part of a larger study. For this work, our interest was only in conservative participants, thus participants who rated themselves above the median on a 9-point conservatism scale at pre-screening were eligible for inclusion in all presented analyses (N = 124).

There was considerable attrition from pre-test screening to post-test; therefore, the final sample of conservatives included in this study consisted of 46 self-identified conservatives ranging in age from 18 to 55 years old with a median age of 27. Given the unique nature of the urban institution, the sample was quite diverse in terms of conservatism, creating a sample that is somewhat different demographically from mainstream conservatives. Thirty-one females and 15 males completed the post-screening: 15 were African American, 15 were Latino(a), 13 were Caucasian, 2 were Asian/Pacific Islander, and 1 participant identified as Multiracial. Participants who completed both the pre- and the post-screening instruments did not differ from their counterparts who only completed the pre-screening in ethnicity or gender (p's>.05); however, they did differ in age such that participants who did not complete the post-screening and manipulations had a median age of 22, while those who completed both pre- and post-screening had a median age of 27, t(122) = 2.62, p = .01. Age was not a significant covariate in any of the subsequent hypothesis tests; therefore, it was not included as a covariate in any of the reported analyses.

Participants were recruited through a university research participation pool and received partial course credit for participation. After signing up for participation via the online recruiting site, participants were automatically linked to an online survey (this process was the same for pre- and post-screenings). At pre-screening, participants completed an extensive online survey, which included a pre-screening measure of conservatism. Approximately one-week after completing the pre-screening, participants completed a second online survey, which included the manipulations and dependent measures. First, participants read a transcript of a Tea Party leader's speech that advocated fiscal conservatism and a return to “God in American politics” (see Appendix). While this speech transcript was taken directly from a Tea Party website, the name of the leader was left anonymous to prevent biasing participants familiar with the Tea Party movement. The comparative context was manipulated such that participants were told that the speech was delivered to a large “Convention of Democrats” (intergroup) or to large “Convention of Republicans” (intragroup). Participants were then primed with either high or low uncertainty and then completed the dependent measures as well as survey items not included in this study.

Procedure and design

Conservatism. At pre-screening, participants completed a three-item measure of conservatism (1 very liberal, 9 very conservative), α = .90 (e.g., “How would you describe your general political views?”). Responses to the composite of these items were split at their median (5) so that we could address our hypotheses, which are specifically directed toward conservatives for whom the Tea Party can be viewed as part of the ingroup “Conservatives.”

Tea Party speech. One week after completing the pre-screening measurements, participants read a speech given by a leader of the Tea Party, in which the leader advocated a “return” to limiting government and social programs and a “return” to God in American politics (see Appendix).

Context manipulation. Participants were randomly assigned to a context condition in which we led them to believe that the speech that they read had been delivered to a group of Republicans (thus forming an intragroup context for conservatives) or to a group of Democrats (forming an intergroup context for conservatives).

Uncertainty prime. Next, participants were randomly assigned to a high or a low uncertainty prime adapted from previous research by Hogg and his associates (Hogg, 2006; Hohman & Hogg, 2011; Hohman, Hogg, & Bligh, 2010). The high-uncertainty prime asked participants to think about and list three things that made them feel uncertain about their lives and the low-uncertainty prime asked participants to think about and list three things that made them feel confident about their lives.

Group support of message. Participants completed two items that assessed their perceptions of how much their groups (political party) supported the message in the speech: “This speech is representative of the way in which my political party views American politics”; “In general, my political party would agree with the speakers' overall message in this speech,” (1 strongly disagree, 9 strongly agree), α = .93; r(45) = .91, p < .001.

Agreement with message. One item assessed participants' agreement with the statement: “Overall, I agree with the message in this speech” (1 strongly disagree, 9 strongly agree).

Importance of message to America. One item assessed participant's agreement with the statement: “This speech highlights the most important issues currently facing America” (1 strongly disagree, 9 strongly agree).

The three dependent measures were presented to participants directly after the experimental manipulations and as these items comprised part of a larger survey, they were followed by several variables not addressed in this work. All demographic data (e.g., age, sex, ethnicity) were collected at the conclusion of the pre-screening measurements.

Results

The design of this experiment includes two manipulated independent variables across which participants were randomly assigned: uncertainty (low vs. high) and context (intragroup vs. intergroup). The key dependent variables were group support of message (group support), agreement with message (message agreement), and importance of message to America (American importance). Reliabilities, means, SDs, and inter-correlation of all variables are shown in Table 1. While the three dependent measures were correlated with one another, we chose to analyze and report the results separately in accordance with our a priori hypotheses. These measures, based on face validity, comprise theoretically different constructs: a third-party ingroup effect, a measure of attitude specifically directed toward the message, and a measure of attitude directed toward nation.

Table 1 Reliabilities, means, SDs, and inter-correlation of variables

Variable	α	M	SD	2	3	4	5
1. Uncertainty prime	–	1.52	.51	− .05	.18	.08	.09
2. Context prime	–	1.52	.51	–	− .002	.12	.09
3. Group support of message	.93	5.34	2.02		–	.76^*^*	.71^*^*
4. Agreement with message	–	5.80	2.13			–	.67^*^*
5. Importance of message to America	–	5.24	2.21				–

Means, standard deviations, reliabilities, and correlations of variables presented here (post-screening) include only participants high in conservatism (N = 47). Means can take values between 1 and 9, with 9 indicating more of the property described, except for the binary uncertainty prime (1 low uncertainty, 2 high uncertainty) and the context prime (1 Republican/intragroup, 2 Democrat/intergroup).

**p < .01.

Group support of message

There were no main effects of either uncertainty or context on participants' perceptions of their political group's support of the message, (p = .23; p = .90, respectively), but the interaction between uncertainty and context was significant, F(1, 42) = 13.53, p = .001. Under low uncertainty, participants in the intragroup context condition (M = 6.05, SD = 1.21) perceived their political group as more supportive of the message than participants in the intergroup condition (M = 4.04, SD = 1.95), F(1, 42) = 6.91, p = .012. In line with our predictions, under high uncertainty, participants in the intergroup context condition (M = 6.62, SD = 1.42) perceived more support for the message than participants in the intragroup context condition (M = 4.75, SD = 2.27), F(1, 42), p = .014. Moreover, participants in the intergroup context perceived more group support for the message under high than low uncertainty, F(1, 42) = 12.57, p = .001 (see Figure 1). Participants in the intragroup context reported marginally more perceived group support under low than high uncertainty, F(1, 42) = 2.89, p = .09.

Figure 1 Perceptions of group's support of message, as a function of uncertainty and message context.

Agreement with message

Once again, there were no main effects of either uncertainty or context on the dependent variable, message agreement (p = .64; p = .46, respectively); however, the interaction between uncertainty and context approached statistical significance, F(1, 42) = 3.29, p = .08. In neither the intra- nor the intergroup context conditions did participants differ in message support as a result of uncertainty (p = .36; p = .11, respectively). In support of our predictions, it was only under high uncertainty that participants in the intergroup context condition (M = 6.75, SD = 1.29) agreed more with the message than participants in the intragroup context condition (M = 5.17, SD = 2.72); F(1, 42) = 3.42, p = .07 (see Figure 2). The simple main effect for context was not significant under low uncertainty (p = .46).

Figure 2 Agreement with the message, as a function of uncertainty and message context.

Importance of message to America

Similar to the previous analyses, there were no main effects of either uncertainty or context on the dependent variable, message importance, (p = .64; p = .64, respectively); and the interaction between uncertainty and context was significant, F(1, 42) = 5.79, p = .02. Under low uncertainty, participants did not differ based on their assignment to context condition in their perceptions of message importance (p = .20). In line with our predictions, under high uncertainty, participants in the intergroup context condition (M = 6.33, SD = 1.67) believed the message was more important than participants in the intragroup context condition (M = 4.50, SD = 2.28), F(1, 42) = 4.44, p = .04. Moreover, participants in the intergroup context believed the message was more important under high (M = 6.33, SD = 1.67) than low (M = 4.50, SD = 2.54) uncertainty, F(1, 42) = 4.44, p = .04 (see Figure 3). The simple main effect for the intragroup context was not significant (p = .20).

Figure 3 Importance of message to America, as a function of uncertainty and message context.

Additional analyses

In addition to our primary hypothesis tests, we also examined the data to explore whether these effects differed among liberals and conservatives. We did not have specific hypotheses for liberal participants, as our primary goal was to test this model in a group that viewed the Tea Party as a wing of their ingroup; however, such findings give us an interesting view of how the message of an outgroup might impact liberals and allows for a comparison of results along conservatism as a continuous variable.

We tested a model that was similar to the primary hypothesis tests, using the full sample for each of the above dependent variables in which we included liberal participants, keeping the conservatism variable continuous. Using three regression analyses, we predicted each dependent variable with conservatism, uncertainty, context, the two-way interactions, and the three-way interaction. All standardized and unstandardized regression coefficients are reported in Table 2.

Table 2 Summary of hierarchical regression analyses predicting group support, message agreement, and message importance (N = 90)

	Group support			Message agreement			Message importance
Variable	B	SE B	β	B	SE B	β	B	SE B	β
Model 1
Context (Cont)	− .36	− .46	− .08	− .04	− .49	− .01	.10	.48	.20
Conservatism (Cons)	.54**	.13	.40**	.51**	.12	.37**	.48*	.14	.35*
Uncertainty (Unc)	− .02	.47	− .01	− .52	.49	− .10	− .49	.49	− .10
Model 2
Cont × Cons	.33	.26	.12	.34	.27	.12	.39	.26	.14
Cont × Unc	2.15	.90	.23*	1.11	.97	.11	2.00*	.93	.21*
Cons × Unc	.32	.25	.12	.59*	.27	.21*	.57*	.26	.21*
Model 3
Cont × Cons × Unc	− .01	− .03	− .01	.42	.55	.07	.17	.53	.03

Table 2 only reports the relevant effects as they appear in each model (not all effects).

*p < .05.

**p < .001.

Across all analyses, conservatism was a significant predictor of each dependent variable, regardless of experimental condition, with people higher in conservatism indicating that their group was more likely to support the message, indicating greater importance of the message to American, and demonstrating more personal support for the message than people higher in liberalism. Conservatism was the only significant main effect for this set of analyses. While none of the three-way interactions reached significance across the three dependent variables, there were some statistically significant two-way interactions. Uncertainty and context interacted to predict party support, such that under high uncertainty, party support was greater when the speech was given to Democrats, while under low uncertainty, party support was greater when the speech was given to Republicans. This two-way interaction is difficult to interpret in isolation from conservatism as a predictor, because for people high in conservatism, the speech given to Republicans formed an intragroup context and the speech given to Democrats formed an intergroup context.

Conservatism and uncertainty interacted to predict message agreement such that in the high uncertainty condition, as conservatism increased, agreement with the message increases. In the low-uncertainty condition, conservatism appears unrelated to message agreement.

Uncertainty and context interacted to predict importance of message such that under high uncertainty people deemed the views espoused in the message more important for America when the speech was given to Democrats. Under low uncertainty, participants deemed the message more important to America when it was given to Republicans. Once again, this finding is difficult to elucidate in isolation from participants' reported conservatism. Conservatism and uncertainty also interacted to predict message importance such that under high uncertainty, as conservatism increased, message importance increased. Under low uncertainty, conservatism appears unrelated to importance.

Discussion

The results of this experiment provide empirical evidence for a shift in political attitudes after exposure to an extreme message from a numerically small political group. In line with research on polarizing prototypes (e.g., Abrams et al., 1990; David & Turner, 1999), our results showed that this extreme Tea Party message is most effective in an intergroup context and when message targets feel uncertain (Hogg, 2012). Research on uncertainty-identity theory demonstrates that when faced with instances of uncertainty, people tend to identify with entitative groups, some of which may also espouse an extreme ideology (Hogg, Meehan, & Farquharson, 2010). Our results provide evidence that when people are uncertain, they not only support the relatively extreme message of an ingroup source, but they also believe that others similar to them have experienced this shift (e.g., other people in their political parties). Interestingly, only in instances of an intergroup context when such an identification with the extreme wing allows the target to polarize away from an outgroup (Abrams et al., 1990) were people willing to support this message.

Taking on the values and attitudes of extreme wings of the ingroup may prove to be a powerful source of uncertainty reduction. Conservatives were most supportive of the message in a comparative intergroup context because polarizing away from an outgroup (Democrats) and toward the message of an extreme wing of their ingroup (the Tea Party) may have allowed for uncertainty reduction. The results of this experiment demonstrate this pattern of polarization consistently across three dependent measures, allowing for a rich assessment of conservatives' political attitudes. Examining both support of the ideology espoused in the message and the importance of such ideology for America allowed for direct measurements of attitudes, while examining political attitudes through a third party (participants' ingroup) allowed us to both assess political attitudes somewhat indirectly (see Duck, Hogg, & Terry, 1995) and examine the broader implications of uncertainty and context on an extreme ingroup faction's ability to impact the overarching group's prototype. Although these three dependent measures were correlated with one another, theoretically, they address different constructs, and viewing them as separate measures allows for comprehensive analysis of political attitudes.

Our results provide evidence that not only is the comparative context important for support of relatively extreme messages and polarization of group prototypes, but so is the overall climate of uncertainty. It is important to keep in mind that our manipulations only examined the effect of a political message from a conservative social group, meaning resulting attitudes and prototypes impacted conservative participants (Hogg & Turner, 1987). We attempted to examine these processes in both liberal and conservative participants in our regression model; however, the results should be interpreted with caution given the small sample size and the amount of variables and interactions included in each model. Moreover, we only examined the extreme message coming from the Tea Party, which is a subgroup of conservatives. We do not currently have data to address the question of whether or not liberals will experience similar attitudes in the face of an extreme ingroup. Taken together with previous findings (e.g., Hogg, Meehan, & Farquharson, 2010) our findings might very well apply to other groups such as the recent “occupy Wall Street” protest groups. This experiment provides the first direct test that demonstrates that in climates of uncertainty, people look to extreme wings of their ingroup and are influenced by the small faction's message if that message can provide a clear and unambiguous identity that is distinct from a relevant outgroup.

Some of the current findings might appear contradictory to other research on political ideology and uncertainty, specifically with respect to the findings of Jost and colleagues (2012), who demonstrate that under high uncertainty, college students are less willing to engage in “system-challenging” (p. 201) protest (i.e., protest against bank bail-outs). In other words, under uncertainty, college students, a fairly liberal group, were unwilling to take the liberal position of protest. We believe that our current findings add an additional layer to this previous research: we highlight the important identity function that groups can serve under uncertainty. Groups that provide a clear identity and distinguish one's ingroup from a relevant outgroup are attractive under uncertainty—people identify with these groups and share in their norms and attitudes. When these groups advocate and are built upon system-challenging ideologies, identification with such groups can mean individuals expressing attitudes that align with these groups' philosophies, in contradistinction to relevant outgroups. In a context that is not explicitly intergroup, such as the case with Jost and colleagues' 2012 work, uncertainty might well be best reduced through adopting a more conservative position that supports the system rather than supporting a philosophy of change, which might increase uncertainty.

The results of our hypothesis tests support an uncertainty-identity analysis of group behavior and social influence (Hogg, 2012), because extreme factions, perhaps by their very nature, allow for uncertainty reduction in particular contexts. The Tea Party and the American conservative movement provide a real-world analysis of such processes—by experimentally manipulating uncertainty and comparative context, this research demonstrates what many U.S. political pundits have suggested: the Tea Party did have the ability to push an agenda of American conservatism that can result in a polarization of American politics.

While the sample size is clearly too small to draw firm conclusions from the regression models examined, these results do offer interesting insight into how liberals might be impacted by the message of an outgroup. It appears that in general, liberals were not influenced by a message from the Tea Party. We cannot use the findings from the significant two-way interactions to suggest that under uncertainty, conservatives might be more influenced by a message than liberals, because these findings are confounded by the fact that for conservatives, the message source was always an ingroup and for liberals the message source was always an outgroup. If people look to their ingroups in times of uncertainty as suggested by uncertainty-identity theory (see Hogg, 2012), then liberals would likely only be influenced by a message from other liberals when they are uncertain.

This experiment is limited by its small sample size and by sample that is demographically unique with respect to the make-up of the broader population of American conservatives (i.e., primarily white; see Pew Research Center, 2011). Although the use of random assignment of participants across an experimental design lends confidence to the validity of the current findings, future work should replicate these findings using a similar experimental design as well as utilizing methodologies that allow for broad generalization of these findings to other social groups.

One additional potential limitation to this work is the operationalization of uncertainty. We refer to the formation of the Tea Party occurring in a time of economic uncertainty. Our theoretical approach to the hypotheses tested here examines uncertainty as self-conceptual uncertainty and we therefore primed uncertainty in accordance with previous work in this area (see Hogg, 2012; Van den Bos, 2009). To the extent that economic uncertainty is self-relevant, the prime that we use will produce similar effects (see Hogg, 2007). Nonetheless, future research might examine how societal level uncertainty and specific uncertainty from the economy impact identification processes.

When people are uncertain, they look for ways to reduce that uncertainty, and extreme wings of their ingroup provide a clear attitudinal and behavioral map to allow for that uncertainty reduction. Particularly in the case of an intergroup context, when people strive to distinguish their important ingroup identity from that of an outgroup, they will turn to the messages of fellow ingroup members who may have otherwise gone unnoticed. In a time when the news is fraught with examples of economic uncertainty and instability, our results highlight how a political social movement may gain influence over moderate members of their political party and polarize group prototypes, perhaps even resulting in social change.

Acknowledgements

The authors thank Fiona Grant for help with editing and William D. Crano for his expertise on attitudes and minority influence.

APPENDIX

Tea Party speech transcript

The following transcript is a speech that a leader in the Tea Party movement gave to a large convention of Democrats (Republicans) to explain the position of the Tea Party and their views of the state of the United States. For copyright issues, we are unable to provide you with the name of the speech giver. Please read this speech very carefully and then respond to the questions that follow.

“Though I would love to be the bearer of an enthusiastic and patriotic message, duty calls me instead to bring a message of a patriot's warning. Our independence is under siege.

As we witness arrogance in action through the presidency of Barack Hussein Obama and a league of progressives that aid him in his hideous push towards an emperor's dream, we, as Americans, must ask ourselves a most serious question in regards to our country. Do we care about our independence and how much?

Our answer will determine the form of action or inaction that we will take during the next 18 months prior to the next presidential election to vote again. We will either reclaim our Republic or seal its demise upon the heaps of American exceptionalism. If you genuinely care, then listen …. and then act. If you're not sure if you should care or not; listen and maybe your eyes will be opened. If you don't care, move out of America!

Those who care must take to arms with the most powerful weaponry known to mankind. It's faster than a speeding bullet, stronger than a locomotive, and it's able to leap not only buildings in a single bound, and especially, runaway governments as well. It is the power of principles that can save a nation and our country needs saving and only by and through the guiding hand of God can we save America!

If we care, we will return to these guiding principles and allow them to once again shape our thoughts, our vision and our actions. Why do I say return? I say that because we as a country have abandoned them and therefore, we have abandoned God.

This did not begin when we voted for Obama; it started when we voted out school prayer; when we legalized abortion; when husbands and wives permit divorce to be an option and give up on their love for each other and marriage vows before God. These are just a few ways we have abandoned God and the principles that He created. And therefore, God has left us to our devices.

We must advocate and work toward an America that respects the right of the individual – that allows us, without taxation for government funded programs, to spend and enjoy the fruit of our hard work! The individual American is not responsible for paying for other people's healthcare! The way that our current government is running limits our freedoms and civil liberties through taxation.

So how do we experience the Power of these Principles manifest upon our country now? Simple but not easy. We repent and return. REPENT!? Yep – In our own way apologize to God for our own arrogance; our own selfishness and our own methods.”