How Elite Partisan Polarization Affects Public Opinion Formation

How Elite Partisan Polarization Affects Public Opinion Formation

James N. Druckman, Northwestern UniversityErik Peterson, Stanford UniversityRune Slothuus, Aarhus University

1

Democracy and Competition• “A competitive political system in which competing leaders and

organizations define the alternatives of public policy in such a way that the public can participate in the decision-making process” (Schattschneider (1960:138).

• Does the tenor and/or intensity of such ELITE competition – which is a defining element of democracy – affect the nature of political decision-making?

2

Democracy and Competition• Our focus the well documented increasingly polarized nature of elite

competition What is the affect of elite Polarization on public opinion formation?

• What is elite polarization?

• The major parties become more distinct from one another.

• The major parties become more internally homogeneous (e.g., McCarty et al. 2006).

3

Polarized Politics• November 3, 2010 Obama: "On [several] issues we're probably going to

have to say there are some issues where there's just too much disagreement to get this done right now.”

4

Democracy and Competition

• How does this affect citizens?

• Debate on causes of elite polarization…

• Debates on whether citizens also polarize – some say yes, some say no, some say they “sort” (i.e., party identification becomes more linked to ideology).

• We ask what we see as a more fundamental question: how does elite polarization affect the policy opinion formation of citizens in general?

5

Outline• Framing and Party Competition (Hypotheses)

• Experimental Test of Elite Partisan Polarization and Framing

• Results

• Conclusion / Implications

6

Framing and Party Competition• Our approach: compare decisions that citizens reach in the presence of

varying levels of perceived elite polarization, all else constant (on average = using an experiment).

• Do opinions formed under conditions of perceived elite polarization differ from ones sans polarization?

7

Framing and Party Competition

• Basis of comparisons (main independent variables): Frame/Argument Strength on an Issue X Party Sponsor X Level of Perceived Polarization.

• Does increased polarization increase the power of partisan cues at the cost of “stronger” frames/arguments?

8

Framing Effects

• Definition: In the course of describing an issue, a speaker’s emphasis on a subset of relevant considerations causes individuals to focus on these considerations when constructing their opinions.

• Example: Politicians, media frame KKK rally request as a:

• free speech issue citizens focus on speech considerations citizens support right to rally

• public safety issue citizens focus on safety considerations citizens oppose right to rally

9

Framing Effects

• Many other examples: • campaign finance (free speech or corruption?)• abortion (rights of mother or rights of unborn child?)• gun control (right to bear arms or public safety?)• affirmative action (reverse discrimination or remedial action?)• welfare policy (humanitarianism or overspending?)• social security (individualism or shared security?)• elections (economy or foreign affairs?)

• A Central Means of Elite Influence on Public Opinion.

• These effects ≠ valence framing effects (Tversky and Kahneman 1979)

10

Extant Work

• Randomly expose individuals to one frame (e.g., free speech) or the other (e.g., public safety) and then immediately measure opinions (often in a laboratory or survey environment).

• Common Result: An individual’s opinion reflects the single frame he/she received (e.g., support or oppose right to rally- investigation of moderators)

11

Extant Work

• Competitive Framing Recent work explores impact of strong versus simultaneous weak frames/arguments (Chong and Druckman, 2007; Hansen, 2007; Druckman, 2010; Aaroe, 2010).

• Strong frames win!

12

What Makes a Frame Strong?• Following psychological work, it is strong when the “audience

perceives it to be compelling” (O’Keefe, 2002: 47, 156).

• What does this mean? Three key attributes:

• Available in ones mind (e.g., does “free speech” make sense when one thinks about hate group rallies? Does “litter in the street”?).

• Accessible does it come to mind, typically via a communication (or chronically)?

• Applicable Relative to other frames/arguments, is it seen as “effective or applicable”?

13

What Makes a Frame Strong?

• Availability and Applicability measured in pre-tests and accessibility manipulated in experiment.

• ALL PERCEPTUAL leaving normative status unclear. (I will return to this.)

14


• Continue to lack understanding of what makes one frame stronger (e.g., tend to be emotional, feed cognitive biases, episodic).

• Leaves many questions open (including normative ones) but this is not our goal here (and again will discuss later).

15


• Instead, we identify STRONG (e.g., relatively applicable) and WEAK frames in contrasting directions on issues (pro or con an issue) via pre-tests with representative respondents.

• Then vary party sponsor and level of polarization

• What matters more perceived argument/frame strength OR party identification and does this vary based on conditions of polarization?

16

Party Competition

• Some prior work on party sponsor and frames (e.g., Slothuus, 2010; Slothuss and de Vreese, 2010; Nicholson, 2011)

Find party sponsor matters, especially on conflicting historical issues, but so do pre-existing values.

• These works do NOT explore competitive framing (i.e., vary frame strength or competition of frames) and do not account for varying partisan environments (i.e., polarization levels).

17

Party Cues

• The same is true on the long-standing literature on party cues – party cues matter but when and how when compared to substantive information?

18

Party Cues

• “In spite of numerous claims about the relative influence of policy attributes and position-taking by party elites [i.e., party endorsements], direct evidence is slight because few studies directly compare the effects of these variables [i.e., arguments versus cues]… “in political debate, cues and frames almost always appear together: Party elites rarely take a position without trying to frame it in a way that will garner support for it.” (Bullock, 2011: 496, 511).

• Bullock finds that party cues have an effect, but do not overwhelm substantive content (also see Arceneaux, 2008; Druckman et al., 2010; Nicholson, 2011).

19

Basic Theory

• Based on motivated reasoning (MR)

• tendency to seek out information that confirms prior beliefs (i.e., a confirmation bias)

• view evidence consistent with prior opinions as stronger or more effective (i.e., a prior attitude effect), and spend more time counter-arguing and dismissing evidence inconsistent with prior opinions

• regardless of objective accuracy (i.e., a disconfirmation bias)

• find your way to your desired conclusion.

20

Basic Theory

• E.g., Democrats (Republicans) may view economic conditions favorably during a Democratic (Republican) administration even if they would view the same conditions negatively if Republicans (Democrats) ruled (e.g., Goren et al., 2009; Huber and Gerber, 2010; Kahan et al., 2012; etc.).

• Note party is not only manifestation of motivated reasoning; this is just one source – can also result from strong issue stances (e.g., Kunda, 2001).

21

Basic Theory

• Hence, all else constant, partisans will view their party’s frame as more effective than a frame not sponsored by their party or a frame sponsored by the other party.

• This theory also suggests that partisans will more likely be moved by their own party’s frame, regardless of the frame’s strength.

• BUT

22

What Accentuates MR?

• Various moderators motivation, prior opinion strength, sophistication.

• Yet, most political scientists, assume MR is the norm (e.g., Taber and Lodge 2006: 767) BUT scant evidence per se of this (see Druckman 2012).

23

What Accentuates MR?

• One key moderator of partisan motivated reasoning strength of partisan identity (Lavine et al. 2012: chapter 5)

• stronger identities opinions more likely based on partisan priors.

• “Partisan ambivalence undercuts judgmental confidence that citizens typically derive from partisan cues….they turn away from these perceptual anchors and pay more attention to the particulars” (Lavine et al., 2012).

24

What Alters Partisan Identity?

• Answer Elite Polarization.

• As parties polarize (i.e., more conflict), partisans become less ambivalent about party identity (e.g., Levendusky, 2009; Slothuss and de Vreese, 2010; Lelkes, Iyengar, and Sood, n.d.).

• “When partisan elites debate an issue and the news media cover it, partisan predispositions are activated in the minds of citizens and subsequently constrain their policy preferences” (Dancey and Goren, 2010: 686).

25

What Alters Partisan Identity?

• “In the American political system, one’s political identity typically means one’s partisan identity…especially in an era of partisan polarization…. In an environment characterized by intergroup disagreement, the desire to seek difference with the out-group will likely be strong” (Nicholson, 2012: 52, 55).

• Implication • Increased perceived elite polarization • increased partisan identity • increased MR • more power of cues over substance.

26

What Can Strong Partisan MR Do?

• MR goal is to confirm an opinion they already hold and thus bolstering a prior opinion, lending confidence to the opinion and viewing it as even more important

• When not MR acting against a prior belief, people may become unsure about what to think and thus less confident and view the opinion as less important

• (e.g., Brader, 2006: chapters 4-5; also see Smith, Terry, Crosier, and Duck, 2005: 168, Wyer, 2010 on connection to social identity and importance).

27

What Can Strong Partisan MR Do?

• Thus THREE key variables: Frame strength, party sponsor, perceived level of polarization.

28

Hypotheses

Hypothesis 1 (no cues or polarization): When presented with opposing strong frames (pro vs. con), individuals’ opinions will not be moved by either frame.

When presented with a strong frame on one side (e.g., pro) and a weak frame on the other side (e.g., con), individuals’ opinions, if affected, will be moved only by the strong frame (based on earlier discussed competitive framing work).

Hypothesis 2: When partisans receive a frame sponsored by their party and a conflicting frame sponsored by the other party, regardless of each frames’ strength, they will view their own party’s frame as more effective and the other party’s frame as less effective (based on basic MR).

29

Hypotheses

Hypothesis 3: When partisans receive a frame sponsored by their party and a conflicting frame, regardless of each frames’ strength, they will be more likely to move in the direction of their party’s frame than in the direction of the other party’s frame (based on basic MR).

Hypothesis 4: In a polarized environment, when partisans receive a frame, sponsored by their party and a conflicting frame sponsored by the other party, regardless of each frames’ strength, they will view their own party’s frame as more effective and the other party’s frame as less effective

– to a greater extent than they do in a non-polarized environment (and/or a non-party cue environment- based on heightened partisan identity and stronger MR).

30

Hypotheses

Hypothesis 5: In a polarized environment, when partisans receive a frame, sponsored by their party and a conflicting frame sponsored by the other party, regardless of each frames’ strength, they will be more likely to move in the direction of their party’s frame than in the direction of the other party’s frame

– to a greater extent than they do in a non-polarized environment (and/or a non-party cue environment- based on heightened partisan identity and stronger MR).

Hypothesis 6: Partisans will view their opinions as increasingly important when receiving a frame with their partisan sponsor (versus a frame without their partisan sponsor), and even more so in polarized conditions (based on importance added to greater availability of consistent information).

31

Experiment• Two survey experiments with a representative US sample in the Spring of

2011.

• We had previously recorded partisanship in the summer of 2010 with standard 7-point question; counted leaners as partisans, as is typical.

• We excluded pure independents (no clear results); also is typical (e.g., Levendusky, 2010; Bullock, 2010). 53% Dems; 47% Reps.

32

Policy Issues In Experiment• Energy policy drilling for oil and gas.

• Immigration policy DREAM Act (Development, Relief, Education for Alien Minors Act).

33

Policy Issues In Experiment• Energy policy drilling for oil and gas.

• Immigration policy DREAM Act (Development, Relief, Education for Alien Minors Act).

• Why these issues?

1. Both received attention prior to our study and were relevant (e.g., Obama was going to allow drilling until Deep Horizon Spill on April 10, 2010).

• DREAM Act was first proposed in U.S. Senate in 2001 and many state legislatures – provides pathway to citizenship for undocumented immigrants entering US before age 16.

34

Policy Issues In Experiment

2. While opinions are relevant, not overly crystallized and somewhat conflicted due to competing considerations (e.g., economics versus security)

35

Policy Issues In Experiment3. The parties do NOT hold consistently dramatically different positions, thereby allowing us to vary the perceived level of elite partisan polarization (e.g. Levendusky 2010, Nicholson 2010).

• Republicans tend to favor Drilling and Democrats tend to favor the DREAM Act but on neither is the divide stark (e.g., VA Dem. Warner supports drilling geography matters), various versions of the DREAM Act have been co-sponsored and supported by both parties

36

Policy Issues In Experiment

• We recognize some partisan position stereotypes on these issues but in this initial foray, we opted to use issues on which polarization could be manipulated, as will be clear; we also do not contradict stereotypes in our design.

37

Policy Frames• Identified relevant frames/arguments via content analysis and prior

work choose seven often used frames.

• Pre-tested with 138 non-students for: direction (pro/con) and strength (availability and applicability). As in prior work:

• Availability listing ideas that come to mind when issue is described.

• Applicability asked on a 7-point scale to rate how effective.

• See paper for list of frames tested…

38

Policy Frames

39

Drilling Frames Supportive (Pro) Opposed (Con) Strong Economic Benefits Worker and Maritime

Life Weak Technological

Developments Regulation

DREAM Act Frames Supportive (Pro) Opposed (Con) Strong Beneficiaries Overburdening the

System Weak Public Support Politics

• On both issues, the pro frames are sig. greater than con (but not each other) and strong are sig. greater than weak (but not each other) in terms of availability and applicability.

Partisan Cues

• How do we operationalization partisan cues/sponsors? We follow Levendusky (2010):

• We never pit a frame with a party cue against one without a cue (does not prevent testing hypotheses) BUT we do have conditions sans party cues.

40

Partisan Cues

• We have parties single positions across all conditions.

• That is, the Democrats always oppose drilling and endorse the DREAM Act – albeit using different frames

• While the Republicans always do the reverse. This is realistic given the parties’ “typical” positions (also see Nicholson, 2012), despite aforementioned universal agreement. This makes the number of conditions feasible and does not prevent testing our hypotheses.

41

Polarization

• How do we operationalize polarization? (also see Levendusky, 2010; although, we rely on text only for realism):

• Three Levels:

• 1. Non-Polarized condition: “the partisan divide is not stark as the parties are not too far apart… members of each party can be found on both sides of the issue.”

• 2. Polarized condition: “the partisan divide is stark as the parties are far apart… most members of each party are on the same side as the rest of their party.”

• 3. No manipulation of polarization.

42

Partisan Cues and Polarization

43

Table 2 Experimental Conditions

Pro Strong Frame Con Strong Frame

Pro Strong Frame Con Weak Frame

Pro Weak Frame Con Strong Frame

Pro Weak Frame Con Weak Frame

No-Party Endorsements (2) N (drilling) = 48 N (DREAM) = 46

(3) N = 55 N = 55

(4) N = 51 N = 49

(5) N = 46 N = 50

Non-Polarized, Party Endorsements For drilling Rep = Pro Dem = Con For DREAM Rep = Con Dem = Pro

(6) N = 51 N = 51

(7) N = 49 N = 44

(8) N = 50 N = 56

(9) N = 51 N = 50

Polarized. Party Endorsements For drilling Rep = Pro Dem = Con For DREAM Rep = Con Dem = Pro

(10) N = 54 N = 52

(11) N = 49 N = 52

(12) N = 45 N = 50

(13) N = 50 N = 44

Note: Ns vary across issues due to non-response. Both experiments included a control condition (1) (N = 47) which received only a brief description of each policy before answering the dependent variables.

Example Vignette (polarized party endorsements, two strong frames)There has been a lot of recent discussion about whether to allow drilling for oil and gas off the Atlantic Coast and in the eastern Gulf of Mexico. Republicans in Congress tend to favor drilling and Democrats in Congress tend to oppose drilling. Moreover, the partisan divide is stark as the parties are far apart. Also, not only do Republicans tend to be in favor and Democrats opposed, but most members of each party are on the same side as the rest of their party. The main argument for those in favor of drilling is that drilling increases our oil supply, which leads to lower gas prices. It also generates employment opportunities and development. The main argument for those opposed to drilling is that workers are required to learn new skills in order to protect themselves against the dangers of drilling. Marine life also must adapt to survive in the face of site construction and drilling.

44

Example Vignette (non-polarized endorsements, strong pro, weak con)

There has been a lot of recent discussion about whether to allow drilling for oil and gas off the Atlantic Coast and in the eastern Gulf of Mexico. Republicans in Congress tend to favor drilling and Democrats in Congress tend to oppose drilling. However, the partisan divide is not stark as the parties are not too far apart. Also, while Republicans tend to be in favor and Democrats opposed, members of each party can be found on both sides of the issue. The main argument for those in favor of drilling is that drilling increases our oil supply, which leads to lower gas prices. It also generates employment opportunities and development. The main argument for those opposed to drilling is that government regulators oversee the drilling. These regulatory agencies recently have expressed being overwhelmed by oversight tasks.

45

Minor Details• Stimuli in conditions 2-5 made no reference to parties and left out the

paragraph on party endorsements.

• We purposely avoided overly blunt party cues by de-coupling the cues from the arguments. Hence, we presented party cues and frames as potentially competing information in order to illuminate what information people utilize.

46

Minor Details

• We assigned all participants to conditions on both issues, and they always received information about drilling first.

• Participants were also assigned to the same conditions on each issue as we worried it would seem disorienting to vary the extent of polarization across issues.

• Dynamics are very similar across issues, and hence there is little reason to assume/believe that partisan stimulus accumulating over time reinforced its effects (i.e., if that would have been the case, partisan differences should increase in magnitude to the second experiment).

47

Measures• Overall Support (hyp. 1, 3, 5); Effectiveness (hyp. 2 and 4); opinion strength

(hyp. 6).

• Overall support “Given this information, to what extent do you oppose or support drilling for oil and gas off the Atlantic Coast and in the eastern Gulf of Mexico?,” with answers on a fully labeled 1 to 7 scale ranging from strongly oppose to strongly support.

• Effectiveness “…how effective or ineffective did you find the main argument opposed to drilling [the DREAM Act]?”, with response options offered on a 1 to 7 scale ranging from completely ineffective to completely effective.

48

Measures

• Opinion strength “How important to you is your opinion about drilling [towards the DREAM Act] (e.g., how strongly do you feel about your opinion)?” (1-7; see Visser et al., 2006).

49

Presenting the Results• As with others (e.g., Taber and Lodge 2006; Druckman and Bolsen 2011)

asking overall opinion and argument effectiveness questions in tandem runs the risk of contamination. We thus rely on our pre-test results to offer additional evidence on frame strength and view it as an area of future work.

• All results robust to controls.

• We included a manipulation check to ensure that participants registered the parties’ endorsements and recalled the frame; over 85% did both across conditions.

• Present overall support results for the DREAM Act by looking at the percentage change in opinion, by condition, relative to the control group.

50

Overall Drilling Support w/No Cues or Polarization

51

2.2%

18.8%***

-10.8%**

5.2%4.9%

14.0%**

-10.4%*

3.5%

-30%

-20%

-10%

0%

10%

20%

30%

Pro-Stg -Con-Stg (2)

Pro-Stg -Con-Wk (3)

Pro-Wk -Con-Stg (4)

Pro-Wk -Con-Wk (5)

Shift

in S

uppo

rt(r

elati

ve to

con

trol

)

Party Condition

Figure 1a: Drilling Support, No Party Cues

DemocraticRespondents

RepublicanRespondents***p≤.01; **p≤.05; *p≤.10, one-tailed

• In support of hyp. 1, sans party cues or polarization, frame strength dominates (and frames of equal strength cancel), regardless of PID. The effects are large (e.g., 25% swing).

Overall Drilling Support w/ cues and Low Polarization

52

-5.7%

15.2%**

-10.1%*

-2.8%

10.6%**14.4%**

-10.7%*

10.2%*

-30%

-20%

-10%

0%

10%

20%

30%


Pro-Stg -Con-Wk (7)

Pro-Wk -Con-Stg (8)

Pro-Wk -Con-Wk (9)

Shift

in S

uppo

rt(r

elati

ve to

con

trol

)

Party Condition

Figure 2a: Drilling Support, Non-Polarized Party Cues (Republicans Pro, Democrats Con)



• Mixed support for hyp. 3 strong frames outperform weak frames regardless of party cues party cues do NOT dominate (Bullock). BUT when frames are of equal strength, party cues drive opinion lexicographical reasoning where substantive information is first piece processed.

Overall Drilling Support w/ cues and HIGH Polarization

53

10.6%**14.4%**

-10.7%*

10.2%*

-30%

-20%

-10%

0%

10%

20%

30%


Pro-Stg -Con-Wk (7)

Pro-Wk -Con-Stg (8)

Pro-Wk -Con-Wk (9)

Shift

in S

uppo

rt(r

elati

ve to

cont

rol)

Party Condition

Figure 2a: Drilling Support, Non-Polarized Party Cues (Republicans Pro, Democrats Con)


• Support for hyp. 5 with high polarization party cues dominate over substance, regardless of the frame strength. Polarization appears to heighten identity leading to increased MR (e.g., 28% swings).

Overall DREAM Support w/No Cues or Polarization

54

2.2%

18.8%***

-10.8%**

5.2%4.9%

14.0%**

-10.4%*

3.5%

-30%

-20%

-10%

0%

10%

20%

30%


Pro-Stg -Con-Wk (3)

Pro-Wk -Con-Stg (4)

Pro-Wk -Con-Wk (5)

Shift

in S

uppo

rt(r

elati

ve to

con

trol

)

Party Condition

Figure 1a: Drilling Support, No Party Cues



• In support of hyp. 1, sans party cues or polarization, frame strength dominates (and frames of equal strength cancel), regardless of PID. The effects are large (e.g., 25% swing).

Overall DREAM Support w/ cues and Low Polarization

55

11.1%

26.1%***

-16.1%**

17.7%**

-11.8%*

12.6%*

-19.6%***

-8.5%

-30%

-20%

-10%

0%

10%

20%

30%


Pro-Stg -Con-Wk (7)

Pro-Wk -Con-Stg (8)

Pro-Wk -Con-Wk (9)

Shift

in S

uppo

rt(r

elati

ve to

cont

rol)

Party Condition

Figure 2b: DREAM Act Support, Non-Polarized Party Cues (Democrats Pro, Republicans Con)



• Mixed support for hyp. 3 strong frames outperform weak frames regardless of party cues party cues do NOT dominate (Bullock 2011). BUT when frames are of equal strength, party cues drive opinion lexicographical reasoning where substantive information is processed first.

Overall DREAM w/ cues and HIGH Polarization

56

17.2%** 16.3%** 15.9%** 17.6%**

-18.4%*** -19.2%***-21.7%***

-24.3%***

-30%

-20%

-10%

0%

10%

20%

30%


Pro-Stg -Con-Wk (11)

Pro-Wk -Con-Stg (12)

Pro-Wk -Con-Wk (13)

Shift

in S

uppo

rt(r

elati

ve to

cont

rol)

Party Condition

Figure 3b: DREAM Act Support, Polarized Party Cues(Democrats Pro, Republicans Con)



• Support for hyp. 5 with high polarization party cues dominate over substance, regardless of the frame strength. Polarization appears to heighten identity leading to increased MR (e.g., 28% swings).

Summary• When presented with opposing frames of the same strength (with no party

endorsements), individuals’ opinions are not affected.

• When presented with opposing frames of differing strength (with no party endorsements), individuals’ opinions move only in the direction of the strong frame.

57

Summary• Under conditions of low-polarization AND CUES:

• when presented with opposing frames of different strength (e.g., one strong and one weak), endorsed by different parties, partisans’ opinions move only in the direction of the strong frame regardless of the party endorsements. In this case, substance outweighs partisan cues.

• when presented with opposing frames of similar strength (e.g., both strong or both weak), endorsed by different parties, partisans’ opinions move only in the direction of the frame endorsed by their party.

In this case, party endorsements drive opinions, in the face of arguments that do not differ in strength.

58

Summary• Under conditions of high-polarization AND CUES:

• when presented with opposing frames, regardless of strength, partisans’ opinions move only in the direction of the frame endorsed by their party. In this case, party endorsements drive opinions, regardless of argument strength – that is, even if the other party’s argument is stronger.

• Thus, parties do not always rule public opinion. However, substance becomes irrelevant and partisan cues win out when elites are perceived as polarized.

59

Summary

• Perceived Polarization clearly plays a role in shaping how policy opinion formation works: it stimulates greater motivated reasoning.

Note: proper test for hyp. 5 compares low and high polarization condition changes and we largely find sig. changes across comparing just those conditions (always at least very close to sig.).

60

Frame Effectiveness• Recall we predict that partisan cues should affect assessments of frame

strength (hyp. 4 and 5).

• We do not explore if perceptions of strength mediate opinions (see Bullock and Ha, 2010).

• Simple summary of results (detailed results similar to overall opinion results and can present the tables with details in Q&A if prefer).

• Recall question asked to rate “effectiveness” of the frame in making an argument.

61

Frame Effectiveness Summary (on both issues)

• With no party cues/polarization strong frames are nearly always seen as significantly more effective than weak frames (and equal strength pro/con frames as equally effective), regardless of party. Analogous to overall opinion result.

62


• With party cues and low polarization when the frames differ in strength, partisans nearly always ignore the cue and rate the stronger frame as more effective regardless of party.

• When frames are equal strength and low polarization, there is a stronger tendency to follow party cues more MR (similar to overall opinion).

63


• With party cues and high polarization partisans always evaluate the frames endorsed by their party as significantly stronger, regardless of issue or party. We see a complete flip in effectiveness evaluations compared to no cue conditions

• This is true for both pro and con frames. (Comparisons between low and high polarization conditions = formal test of hyp 5 largely supportive but not completely supportive results).

• Again, specific results available during Q&A.

64

Opinion Importance• Hypothesis 6: Partisans will view their opinions as increasingly important

when receiving a frame with their partisan sponsor (versus a frame without their partisan sponsor), and even more so in polarized conditions (based on greater availability of consistent information).

• 7-point scale asking for the importance attached to each attitude.

65

Opinion Importance

• Use control group as baseline – does importance increase when partisan cues are added? Does it increase even more under conditions of polarization?

• Present results by exploring % change relative to the control for each condition, partisan average.

66

Drilling Opinion Importance

67

Table 5a Shift in Attitude Importance of Drilling Opinions

(Relative to Control Group) Frames

(Condition) Cues Democrats Republicans

Pro-Stg - Con-Stg (2) No Parties -2.3 % -4.5 %

Pro-Stg - Con-Wk (3) No Parties 3.2 % 7.7 %

Pro-Wk - Con-Stg (4) No Parties -1.0 % 0.7 %

Pro-Wk - Con-Wk (5) No Parties -2.4 % -6.3 %

Pro-Stg - Con-Stg (6)

Non-Polarized -0.5 % 14.6 %***

Pro-Stg - Con-Wk (7)

Non-Polarized 3.9 % -4.4 %

Pro-Wk - Con-Stg (8)

Non-Polarized 5.3 % 7.1 %

Pro-Wk - Con-Wk (9)

Non-Polarized 2.1 % 8.8 %**

Pro-Stg - Con-Stg (10) Polarized 11.6 %*** 15.1 %***

Pro-Stg - Con-Wk (11) Polarized 12.9 %*** 16.3 %***

Pro-Wk - Con-Stg (12) Polarized 10.8 %** 13.8 %***

Pro-Wk - Con-Wk (13) Polarized 12.4 %** 13.8 %***

***p≤.01; **p≤.05; *p≤.10 for one-tailed tests (relative to the control group). • Without polarization (even with party cues) we see increased importance in only two

cases. But with polarization, we see dramatic increases in all cases (by over 13%).

DREAM Opinion Importance

68

Table 5b Shift in Attitude Importance of DREAM Act Opinions

(Relative to Control Group) Frames

(Condition) Cues Democrats Republicans

Pro-Stg - Con-Stg (2) No Parties 2.0 % 1.6 %

Pro-Stg - Con-Wk (3) No Parties 0.9 % -4.2 %

Pro-Wk - Con-Stg (4) No Parties -4.1 % 7.6 %

Pro-Wk - Con-Wk (5) No Parties 6.8 % 7.6 %






Non-Polarized -9.0 %* 7.1 %

Pro-Wk - Con-Wk (9)


Pro-Stg - Con-Stg (10) Polarized 16.0 %** 18.1 %***

Pro-Stg - Con-Wk (11) Polarized 12.3 %** 17.1 %***

Pro-Wk - Con-Stg (12) Polarized 12.6 %** 17.1 %***

Pro-Wk - Con-Wk (13) Polarized 9.8 %* 12.4 %*

***p≤.01; **p≤.05; *p≤.10 for one-tailed tests (relative to the control group).

• Without polarization (even with party cues) we see no increased importance (although marginal in condition 8 Dems). But with polarization, we see dramatic increases in all cases (by over 12%).

Implications• Clearly, a perceived polarized environment leads individuals to engage in MR

and increases the confidence they have/importance they see in their opinions.

• Thus, not only does elite polarization decrease reliance on substance (otherwise “perceived” as strong), but it affects importance, which in turn has been shown to drive willingness to persuade, take action, over confidence, etc. (Visser et al., 2006).

• This also may generate stability in political partisan identification (Johnson and Fowler, 2011).

69

Conclusions• In the absence of party references, the strength of frames/arguments shape

opinions.

• Frame strength (i.e., perceived substance) continues to shape opinions even with party cues sans polarization and unequal perceived argument strength (Bullock, Nelson).

• But party cues shape opinion much more sans polarization if the competing arguments are of equal strength (i.e., party sponsor is the second piece of information process).

70

Conclusions

• WITH polarization, the order of priorities shifts and partisan identity presumably leads to intense motivated reasoning and party cues largely shape opinions regardless of frame strength.

• They also become more confident in their opinions .

• OVERALL perceived elite polarization fundamentally changes the manner in which citizens make decisions.

71

Implications• Past work suggests MR has negative implications for opinion quality since it

biases interpretation of information.

• Lavine and his colleagues (n.d.) state that MR:

“raises deeply troubling questions about political representation… how can an electorate possibly reward or punish an incumbent party if it holds

grossly distorted views of political conditions? And how can it elect leaders who will pursue desired policy reform in the face of widespread

misperception about where leaders stand, what the policy status quo is, and what the central elements and likely consequences of proposed reform are?” (chapter 5: 6; also see Jerit, 2009).

72

Implications

• Moreover, partisan motived reasoning, as is made clear by Chong and Druckman (2010), Druckman and Leeper (n.d.b.), and Druckman et. al. (2012), can lead to dogmatic adherence to a prior opinion to the point of extreme inflexibility and intolerance. ….

73

Implications• We are sympathetic to this BUT:

• This presumes frames/arguments PERCEIVED to be stronger have more desirable normative qualities.

• Do they? They tend to be stronger when episodic, emotional, etc. Is this desirable?

• Normative and empirical theorists have scant to say on what makes for a good basis for a “reasoned opinion” other than copious information (sometimes of little direct relevance to the opinion) (Lupia, 2006).

• How do people arrive at perceptions of elite polarization?74

Implications

• An implication of our finding is increased elite polarization = increase opinion polarization on specific issue opinions. This contrasts with Levendusky who points out that this at least leads to more opinion constraint a la Converse (1964). Perhaps this is better (more ideological coherence).

• Perhaps using party cues is normatively desirable and perhaps they are more predictive? (see Nicholson and Heit, 2012, who show that people understand what parties represent)

75

Implications

• The larger unanswered question is “what is a ‘better’ or ‘worse’ opinion?” We don’t know.

• For years, many theorists have assumed opinions are fixed. Clearly they are not and are influenced by elites. When is this good or bad?

• We need increased dialogue to explore this and consequently what polarization means for opinion formation from a normative perspective (and ultimately if strong parties are good or bad as well!).

76

Implications• Democratic competition is a defining element of democracy but competition

has recently been shown to:

• Lead to recency effects (Chong and Druckman, 2010).

• Lead to motivated search for reinforcing information (MR search) (Druckman et al., 2012).

• May lead to lower quality opinions due to polarization (?).

• Clearly, the timing, nature, and intensity of competition affects preference formation and future work needs to address these political realities.

77

End

78

Dem Drilling Frame Effectiveness Scores

79

Table 3a Democratic Evaluations of Drilling Frames

Frames (Condition) Cues Pro Frame Con Frame Difference N


No Parties 4.26 (std. dev. = 1.72)

4.37 (1.31)

-0.11 (2.69)

27


No Parties 4.68 (1.58)

3.39 (1.62)

1.29*** (2.66)

38



4.89 (1.48)

-1.59*** (2.22)

27

Pro-Wk - Con-Wk (5)


3.27 (1.76)

0.23 (2.34)

26


Non-Polarized

4.15 (1.19)

4.46 (1.30)

-0.31 (2.19)

26


Non-Polarized

4.44 (1.58)

3.74 (1.48)

0.70* (2.54)

27


Non-Polarized

3.73 (1.66)

4.80 (1.54)

-1.07** (2.88)

30

Pro-Wk - Con-Wk (9)

Non-Polarized

3.60 (1.12)

3.68 (1.18)

-0.08 (1.47)

25


Polarized 3.65 (1.50)

4.96 (1.30)

-1.30*** (2.38)

23



4.89 (1.45)

-1.26** (3.18)

19



4.92 (1.29)

-1.40*** (2.08)

25

Pro-Wk - Con-Wk (13)


4.79 (1.57)

-1.31*** (2.16)

29

***p≤.01; **p≤.05; *p≤.10 for one-tailed tests (within-condition differences).

Rep Drilling Frame Effectiveness Scores

80

Table 3b Republican Evaluations of Drilling Frames




4.86 (1.68)

0.57 (2.25)

21



3.12 (1.83)

2.18*** (2.58)

17



4.75 (1.59)

-1.00** (2.78)

24

Pro-Wk - Con-Wk (5)


3.40 (1.39)

0.35 (2.35)

20


Non-Polarized

5.36 (1.32)

3.64 (1.66)

1.72*** (2.46)

25


Non-Polarized

5.23 (1.45)

3.00 (1.15)

2.23*** (1.88)

22


Non-Polarized

3.70 (2.20)

4.60 (1.76)

-0.90* (2.99)

20

Pro-Wk - Con-Wk (9)

Non-Polarized

5.27 (1.28)

3.38 (1.36)

1.88*** (2.21)

26



3.19 (1.85)

2.42*** (3.03)

31



3.07 (1.44)

2.60*** (2.18)

30



2.75 (1.52)

2.60*** (2.66)

20



3.19 (1.63)

1.90*** (2.45)

21


Dem DREAM Frame Effectiveness Scores

81

Table 4a Democratic Evaluations of DREAM Act Frames




4.21 (1.68)

0.38 (2.87)

29



3.35 (1.66)

1.68*** (2.82)

31



4.85 (1.30)

-1.27*** (2.49)

33

Pro-Wk - Con-Wk (5)


3.84 (1.57)

-0.08 (2.52)

25


Non-Polarized

4.74 (1.53)

4.19 (1.51)

0.55 (2.59)

31


Non-Polarized

5.26 (1.05)

2.78 (1.24)

2.48*** (2.11)

23


Non-Polarized

3.67 (1.43)

4.50 (1.32)

-0.83** (1.95)

24

Pro-Wk - Con-Wk (9)

Non-Polarized

4.20 (1.47)

3.60 (1.63)

0.60* (2.39)

30



3.71 (1.73)

1.04** (3.06)

24



3.08 (1.57)

1.88*** (2.97)

26



3.29 (1.71)

1.38*** (2.48)

24



3.09 (1.48)

1.77*** (2.05)

22


Rep DREAM Frame Effectiveness Scores

82

Table 4b Republicans Evaluations of DREAM Act Frames

Frames (Condition) Cues

Pro-Statement

Con-Statement Difference N



4.94 (1.78)

-0.41 (2.24)

17



3.75 (1.39)

0.50 (2.30)

24



5.31 (1.89)

-1.63* (3.69)

16

Pro-Wk - Con-Wk (5)


3.68 (1.46)

-0.16 (2.43)

25


Non-Polarized

4.15 (1.57)

5.15 (1.46)

-1.00* (2.73)

20


Non-Polarized

4.33 (1.71)

3.67 (1.49)

0.67 (2.96)

21


Non-Polarized

3.28 (1.30)

4.94 (1.70)

-1.66*** (2.15)

32

Pro-Wk - Con-Wk (9)

Non-Polarized

3.00 (1.30)

3.95 (1.43)

-0.95** (2.11)

20



5.18 (1.42)

-1.82*** (2.67)

28



4.96 (1.66)

-1.96*** (3.18)

26



4.92 (1.38)

-2.08*** (2.76)

26



5.00 (1.63)

-2.23*** (2.72)

22


Experimental Design: Varying Attributions• “The Energy Act

• overall, was widely supported by Democratic representatives and,

OR

• overall, was widely supported by Republican representatives and,

OR

• overall, was widely supported by representatives from both parties, and,

OR

• overall, was supported by some, but not all, representatives of both parties, and,

83

Experimental Design: Varying Attributions

• … (details of act)…

• Given this information, to what extent do you oppose or support the Act? (7-point scale).

84

Experimental Design: Varying Motivation• Motivated reasoning depends on individual motivation.

• Accuracy motivation increased stakes involved in making a wrong judgment, without increased attractiveness of a particular opinion.

• Directional motivation motivated to arrive at a particular conclusion (e.g., consistent with party).

85

Experimental Design: Varying Attributions

• [When thinking about your opinion,]

• please try to view the policy in an evenhanded way. We may later ask that you justify the reasons for your judgment – that is, why the policy’s content is more or less appealing. (accuracy)

OR

• consider that the bill was passed during a period of divided government where fellow partisans voted together nearly 90% of the time. This was necessary to ensure coherent policy programs. We may later ask you about your party and why you affiliate with it. (directional)

86

Experimental Conditions

• Normative standard accuracy motivation without presence of cues = deliberative opinions. Ideal for some.

87

None Same

Party Different Party

Consensus Bi-partisan

None ?

Depends

Depends

Depends

No change

Directional ?

Increase Support

Decrease Support

Increase Support

No change

Accuracy Normative Standard

No change

No change

No change

No change

Experimental Conditions

• How do people “typically” process information?

• Does accuracy motivation temper partisan effects?

• Does bi-partisan attribution temper partisan effects, regardless of processing approach?

88

None Same

Party Different Party

Consensus Bi-partisan

None ?

Depends

Depends

Depends

No change

Directional ?

Increase Support

Decrease Support

Increase Support

No change

Accuracy Normative Standard

No change

No change

No change

No change

89

No Party,No Motiv.

No Party,Directional

Same Party,No Motiv.

Same Party,Directional

Same Party,Accuracy

Other Party,No Motiv.

Other Party,Directional

Other Party,Accuracy

Consensus,No Motiv.

ConsensusDirectional

ConsensusAccuracy

Bi-Partisan,No Motiv.

Bi-PartisanDirectional

Bi-Partisan,Accuracy

-15%

-10%

-5%

0%

5%

10%

15%

20%

6.43%* 6.82%**7.60%**

13.97%***

-2.46%

-9.92%***-11.55%***

1.98%

8.65%**

10.53%***

3.69%

1.78%

4.14%

0.25%

Support for 2007 Energy Act(All Respondents)

Conditions

Shift

in S

uppo

rt(R

elat

ive

to C

ontro

l)

***p≤.01; **p≤.05; *p≤.10, one-tailed

• ALL comparisons relative to the normative standard of accuracy induced, no sponsor.

• The no party conditions increase support.

• In the party sponsor no motivation & directional conditions motivated reasoning.

• Same party conditions increase support. • Other party conditions decrease support.

• EVEN THOUGH CONTENT DOES NOT CHANGE.90

No Party,No Motiv.




Same Party,Accuracy




Consensus,No Motiv.


ConsensusAccuracy




-15%

-10%

-5%

0%

5%

10%

15%

20%

6.43%* 6.82%**7.60%**

13.97%***

-2.46%

-9.92%***-11.55%***

1.98%

8.65%**

10.53%***

3.69%

1.78%

4.14%

0.25%


Conditions

Shift

in S

uppo

rt(R

elat

ive

to C

ontro

l)

***p≤.01; **p≤.05; *p≤.10, one-tailed

• Consensus conditions = same party conditions = motivated reasoning.

• This type of universal support does not stimulate a focus on content – it is just like receiving an in-party cue support increases regardless of content. 91

No Party,No Motiv.




Same Party,Accuracy




Consensus,No Motiv.


ConsensusAccuracy




-15%

-10%

-5%

0%

5%

10%

15%

20%

6.43%* 6.82%**7.60%**

13.97%***

-2.46%

-9.92%***-11.55%***

1.98%

8.65%**

10.53%***

3.69%

1.78%

4.14%

0.25%


Conditions

Shift

in S

uppo

rt(R

elat

ive

to C

ontro

l)

***p≤.01; **p≤.05; *p≤.10, one-tailed

• The no motivation conditions = directional manipulations in all cases.

• When not prompted, individuals engage in partisan motivated reasoning.92

No Party,No Motiv.




Same Party,Accuracy




Consensus,No Motiv.


ConsensusAccuracy




-15%

-10%

-5%

0%

5%

10%

15%

20%

6.43%* 6.82%**7.60%**

13.97%***

-2.46%

-9.92%***-11.55%***

1.98%

8.65%**

10.53%***

3.69%

1.78%

4.14%

0.25%


Conditions

Shift

in S

uppo

rt(R

elat

ive

to C

ontro

l)

***p≤.01; **p≤.05; *p≤.10, one-tailed

• In every case, the accuracy inducement eliminated motivated reasoning the presence of a partisan sponsor did not change opinions relative to the normative baseline.

• When asked to justify opinions, people consider content. 93

No Party,No Motiv.




Same Party,Accuracy




Consensus,No Motiv.


ConsensusAccuracy




-15%

-10%

-5%

0%

5%

10%

15%

20%

6.43%* 6.82%**7.60%**

13.97%***

-2.46%

-9.92%***-11.55%***

1.98%

8.65%**

10.53%***

3.69%

1.78%

4.14%

0.25%


Conditions

Shift

in S

uppo

rt(R

elat

ive

to C

ontro

l)

***p≤.01; **p≤.05; *p≤.10, one-tailed

• Bi-partisan conditions eliminate motivated reasoning.

• The presence of a bi-partisan (conflict) did not change opinions relative to the normative baseline. Lead to a focus on content. 94

No Party,No Motiv.




Same Party,Accuracy




Consensus,No Motiv.


ConsensusAccuracy




-15%

-10%

-5%

0%

5%

10%

15%

20%

6.43%* 6.82%**7.60%**

13.97%***

-2.46%

-9.92%***-11.55%***

1.98%

8.65%**

10.53%***

3.69%

1.78%

4.14%

0.25%


Conditions

Shift

in S

uppo

rt(R

elat

ive

to C

ontro

l)

***p≤.01; **p≤.05; *p≤.10, one-tailed

Summary• Parties can hamper citizen decision-making, as partisans selective

analyze information based on the partisan source instead of substantive content.

• They accept (reject) information that they would otherwise reject (accept) had they analyzed the content.

• Three Antidotes (reduce partisan bias):

• Increased individual ambivalence about a person’s party.

• Explicit inducements to form “accurate” decisions.

• Bi-partisan or cross-partisan support (that is not universalistic).95

Polarization

• Why polarization likely increases partisan motivated reasoning

• All else constant, bi-partisan (cross-partisan) sponsorship is less likely to occur when parties polarize (as there is less overlap between members).

• Polarization itself tends to strengthen partisan identity and trust in parties (Levendusky 2010, Slothuus and de Vreese 2010).

96

End

97

How Elite Partisan Polarization Affects Public Opinion Formation

Documents

Transcript of How Elite Partisan Polarization Affects Public Opinion Formation