POLI 311

What is an attitude? 

• Main reference point Phil Converse (1964) 

• Attitude = evaluative judegement of some object or some concept 

• Attitudes can vary in terms of: 

  • Direction (for/against), intensity, consistency over time, the knowledge or beliefs supporting it or otherwise linked to it

What is an opinion?

•  expressed version of an attitude; measurable but attitudes are not 

What is a belief system? 

• Configuration of ideas and attutidues in which the elements are bound together by some form of constraint of function interdependence ie what goes with what 

• Eg. Tariffs influence perception of Mark Carney which influences perception of the Liberal Party 

• Beliefs exist bc of how attitudes are tied together 

• Constraint has different forms

  • Logical 

  • Social (bc friends/colleagues/whatever believe this) 

  • Psychological (eg. i was raised to believe this, experiences, personality etc) 

• Pattern of constraints between attitudes = belief system

• Do people thing ideolgoically about politics? (Converse study methods)

• Do people exhibit constrain in their political opinions (ie exhibut belief systems), and does it map onto the standard lib-cons spectrum? 

• Analysis the first waves of the National Election Study, which also included a “panel” component (re-interviewing the same people in each period) 

• Main part involves analyzing opinions about the 2 political parties and their prez candidate in terms of how abstract and ideological people appear to be 

  • Listing things they like about the party 

• Do people thing ideolgoically about politics? (Converse study findings)

• Finds levels of abstraction 

  • Ideologues = people who actively rely on ideology in some abstract sense when it come to political attitudes 

  • Near-Ideologues = mentioned ideologly in some way but not placing much evaluative weight on it 

  • Group interest = evaluated parties based on who the candidates where good and bad for (most ppl fall in this category) 

  • Nature of the Times = some policy considerations but nothing that would qualify at the higher level (eg, war, peace etc) 

  • No Issue Content = evaluation contain no shred of policy signficance (eg like their smile etc) 

• Most people are NOT ideological thinkers 

• Other findings 

  • People’s position on one issues does not predict their position on another issue (r=0.23 on average; low levels of constraint)

  • Even lower constraint across domains (ie foreign policy and domestic policy) 

  • Masses are MUCH less constrained that poliical elites asked similar questions 

  • Overtime knowing a person’s opinion at one point in time generally does not help us predict that same opinion down the road -- people bounce around -- indicates that people might be guessing in interviews 

  • For a lot of people opinions are generated as if random

P.E. Converse, “Assessing the Capacity of Mass Electorates,”

• Where political information is concerned the level is low but the variance is high in information distribution (331) + heterogeneity is underestimated

• Means that we have to asses the electorate as stratified rather than as uniform 

• And doubts that mass publics having trouble giving tables responses to conventional policy issues bc the questions are poorly worded 

J. Krosnick, “Government Policy and Citizen Action: A Study of

Issue Publics in Contemporary America,”

• Discusses the findings of research on cognitive and behavioural consequences of passionate concern ab gov policy 

• Lots of variance in america about the personal importance attached to attitudes on particular policy issues 

• Those whose attitudes are important 

  • More likely to think ab those attitudes, perceive candidates as polarized on the issue, and form preferences based on those attitudes 

  • Resistant to chance and stable over long periods of time 

• American publi c= many small issues publices composed of people passionately concerned about single issues 

S. Feldman and J. Zaller, “The Political Culture of Ambivalence:

Ideological Responses to the Welfare State”

• MAIN ARGUMENT 

  • Study using open ended questions 

  • Most people use values and principles central to the political culture when discussing their preferences 

  • BUT diffusion of diverse values leads to ambivalence in discussions of their positions 

    • Means that these positions may shift depending on the context

  • Public opinion of social welfare defined by internal conflict instead of fixed ideas bc of stimiltaneous commitment to egalitarianism and individualism 

  • Many people do not possess a single attitude towards the welfare state andthis is where instability in public opinion comes from 

• Critical Perspectives (Converse) - Folk Ideologies

• People have constrained belief systems but that they do not necessarily overlap with converse’s liberal-conservative dimension 

• BUT he argues that 

  • No evidence that people were using some other rubric in his own data 

  • Relationships across issues much weaker in the masses than in elites, indicting no real constraint whatsoever 

    • They are not consistent when asking them actual political questions 

  • Instability of political opinions themselves

• Critical Perspectives (Converse) Time and Place 

• Some critics argued that converse may have been right at the time, but that era what unusually placid and relative non-ideological → ppl would have a clearer sense of the stakes and more constraint now 

• Surely the turbulent years of the 60s/70s would have more constraint BUT not really according to replications across countries 

• Issue Publics (krosnick)

• Most people dont care about most political issues enough to make them want to become highly informed; it is also true that most people care deeply about SOME things 

• When making political decisions these people dont need to know about everything about all issues or even think ideologically → just pick someone who care about what they do 

• Simplifies decision making 

  • Easier to obtain info on limited selection of issues 

  • All people need to do is pick a candidate according to the issue they care about 

  • People are more likely to want to obtain information about an issue they care about 

• Converse open to this possibility but left other to flesh it out 

• Attitudinal Ambivalence (zaller) 

• Feel that people DO have predispositions, values etc BUT 

• Not that their heads or empty but they have a bunch of ideas 

• People might appear to generate their issues not at random but because they have à  lots of ideas floating around in their minds and theyare not necessarily constrained or in agreement with eachother 

• When we ask for ppls opinion about something they develop an answer by sampling whatever considerations are at the top of their head 

• Suggtes that the kind of instability converse shows is not happening bc of LACK of attitudes but bc they have LOTS of them and different ones are salient at different times 

• Context dependent (even down to the wording of the question, response options, order of the question etc) 

• Do ideological innocence and non-atittudes matter for individuals?

• Traditional democratic theory states that citizen must be informed (think about what an election is and what it is for) 

• ^ in that context it makes sense that converse’s research was alarming, raise questions about what is the nature of democracy; how can the democratic system thrive? 

  • Snapshot of quiz show knowledge is not what is important but rather ppls ability to learn about an issue as time progresses 

  • Knowledge of and engagement with the issues and their underlying principles is what allows ppl to exercise their voice 

• Individuals first: So what if i dont know anything? 

  • Nobody ever said that non-attitudes applied ti most people or across most issues 

  • Most people know at least something about a narrow set of political issues (according to issue public) 

  • ^^ bc of this they are most motivated and able to develop a more sophisticated view and weight choice based on what candidates say

• Individuals and Heuristics 

• Issue public is part of heuristics 

• Heuristics = information shortcut that people use to make decisions 

  • Unbiased to the degree that it leads people to the same decisions they would have otherwise made had they become informed 

  • Efficient to the degree that it leads people to that conclusion with as little cognitive energy expended as possible 

• All kinds of things are heuristics 

  • Ideology (some ppl support/resist something based on the ideological label)

  • Identities (ie what do ppl i am like support or oppose) 

  • Policy (as in the issue public idea) 

• An efficient and unbiased heretic in practice 

  • NDP v Conservative 

  • Can determine what they think about issues based on the party they belong to

Individual and exogenously imposed constraints

• Heuristics can simplify a political deicsion 

• Simplicity is also imposed by the system as ppl in democratic system are only asked to make simple choice 

  • As a result someone whose belief system is characterized unconstrained and mostly non-attitudunal or someone who is highly ambivaluent can appear soldi cb the system simplifies 

  • Simplifies by 

    • Opinion leaderships designed to infleunce/capture the issue public around a given concern 

    • Values often framed as right vs wrong with little gray areas 

    • Public choices are typically framed as “we” to enocuge ppl to think in terms of social groups 

    • Set of alternative (choice set) is organized well in advance, limited in number, disjuctive, and exhaustive 

    • Eg. Trump saying “i’m on the ballot” → makes the election about him

• Do non-attitudes matter at the systemic level? 

• To what extent should public opinion guide public policy? (core questions for democratic politics) 

• Hinges on models of representation in delegation or direct influence 

• Makes sense to want to keep the public at arm’s length from decision making, after all 

  • Most people lack solid knowledge of most issue, dont think in ideological consisten way etc 

• Can collectives be rational? 

• In order to say how a public largely charactezed by non-attutudes and ideological innocence becomes rational in the aggregate 

• Random collection of non-attitudes provides no clear public opinion system 

• BUT when opinion leaders provide cues…

  • The issue public concerned over that policy will follow 

• Adding compelting opinion leaders and issues publices leads to all kinds of hypotehsese regarding 

  • Net public opinion 

  • Polarization etc 

• An ideologically coherent public? 

• National mood (higher = more liberal) 

  • There is a mood that you can track overtime that seems to respond to realities even if most people dont have strong policy opinions 

• A rational public? US opinion on abortion 

  • Less undulating → relatively consistent over time 

  • People more supportive if it is out of the woman’s control than if it is in their perceived control 

What is the rational public model?

Theory stating that aggregate public opinion is stable, meaningful, and responds rationally to new information and events

• Individual-level noise, non-attitudes, and random measurement errors cancel out through statistical aggregation

• Results in a collective electorate that behaves rationally even if individual citizens lack full political information

• Some criticism of the rational public model 

• Have to be careful about relying to heavily on either individual or aggregate rationality as a supporting pillar of democracy bc 

  • Often want thing that are logically inconsistent 

    • Lack of logical constraint 

    • Eg. federal income tax → most people think they pay too much tax and rest think they apy the right amount 

    • BUT most people want to increase spending on things like schools, healthcare etc 

    • SO then you should be willing to pay more taxes; ie there is no logical constraint 

  • Readily fixate and use irrelevant theories about politics 

• Problems w heuristics 

• People mislabel political objects ie have the wrong belief about what a politician has said or done 

• Misattribute judgements about cause and effect, which matters if their primary heursitric for voting is past performance 

  • Eg. shark attacks are bad,  and if youve seen jaws you knwo they can devastate the local economy 

  • The causes of shark attacks arent random but political 

  • Yet an abnormally large amount of shark attacks in NJ cost Woodrow Wilson the election 

  • In what universe does this even make sense 

  • Same things happens with droughts, pandemics, economic downturn etc 

  • People are bad at assigning cause and effect 

Worldview polarization

• Worldview polarization ^ 

  • Ppl who’s heuristics becomes a fundamental political assumption of others and themselves 

  • Heuristic bias is more than mislabelling and cause and effect 

  • About seeing the world differently when heuristic becomes strongly associated with whole sets of beliefs and behaviours 

  • Eg. Survey about the pandemic 

    • Democrats taking pandemic safety precautions → almost always taking them; even more pronounced when it comes to attitudes (things like getting sick, increasing testing etc) 

    • Republicans are less likely to do these things 

  • Polarized opinion leaderships leads to polarized publics and ultimately policy -- goes hand in ahnd with ppls proclivity to see the world differently depending on party 

  • Instead of one true public opinion we have different ones in different places (ie response to COVID was polarized) 

  • Eg, Presidential election 

    • Trump claiming he won 2020 election 

    • Most republicans did not agree the Biden was the true president in 2021 and viewed as unfair and inaccurate 

    • Overall trends can hide things that are actually deeply polarized 

• Selective perception as a subset of worldview bias 

• SP = heuristics causing bias by actually tainting or warping new information coming into ppls minds AS IT COMES IN 

• Important bc WVP is causally ambiguous -- all kinds of reasons people view the world differently 

  • Ppls reaction to COVID was along party lines but the heuristic may not have caused this - could be underlying factirse 

• In the case of SO ppl are actually seeing the world thru the lens of theat heuristic 

• EXAMPLE - OJ Simpson 

  • Social identity that serves strong heuristic function is ethnic/racial identification 

  • 1994-1995 was most ubiquitous news story 

  • OJ accused of murdering his ex-wife and her friend Ronald Goldman 

  • Black and white people brought different perspectives BUT for diff reasons 

    • Whites = saw profusion of physical evidence and history of spousal abuse → guilty 

    • Black = victim of LAPD conspiracy and systemic racism in the judicial system → victim 

    • Verdict was very polarizing = not guilty 

  • Sure they saw things differently based on their lived experience 

  • But we also see selective perception 

    • Ie evidence that the case was processed differently by each group 

    • White people got more certain of guilt as the trial goes on 

  • SP aand the support for GW Bush 

    • Average tendency goes down BUT democrats rate him way lower than republicans 

    • Ie they are processing the world differently 

  • COVID Concern 

    • Democrats always more concerned than republicans are 

  • EXAM - COMMENT ON THE TREND OVER TIME 

    • Look for evidence of worldview polarization and selective perception; and what is actually happening 

  • Generally ppl are happy when the president is belonging to their party -- see lots of changeover during elections even when the economy or other factors change 

    • People fundamentally changing their mind about the world based on who is president - is this a good measure of democratic competence? 

  • No trump bump in approval after starting iran war 

    • More recent presedneint dont seem to get the same public support as older ones (eg, Bush 89% approval in the gulf war) 

    • ^^ has not gotten more popular

• Difference between SP and WVP 

• SP - lines are not parallel -- signals an event, and the lines either move in tandem or dont 

  • If they dont move in tandem this is evidence of selective perception 

  • Eg. ppl going down in economic ratings in parallel during COVID but then post- election causes divergence, democrats rate economy higher and republicans rate the economy lower 

  • The economy itself didnt really change but the president 

• If the lines say constant and move together this is WVP 

• If all lines followed basically the same there is no SP or WVP

• Relationship between political ideology and political knowledge 

• Recent study across 63 countries - relationship to where ppl placed themselves on the ideological spectrum and how politically knowledgable they appeared 

• Used available measures of factual recall (eg. who is in government, who is the opposition etc) - either you know what’s going on or you dont 

• USA 

  • Ppl in centre have lowest level of recall 

  • Ppl at the extreme ends goes back down 

  • More moderately partisan highest level of recall ---> ppl that are more likely to receive political information but not so devoted that they throw out info that doesnt suit their worldview 

  • Canada looks about the same 

• Argentina 

  • Ppl on the right know less than people on the left 

  • Opposite in Chile

• When attitudes make us uncomfortable 

• Orwell 1984 → concept of doublethink 

  • Goota hold 2 opposing thoughts in your head and make peace with it 

  • Suggest the enormous difficulty of doing this 

• Cognitive consistency and dissonance 

  • Cognitive dissonance = inconsititences among a person’s attitudes that can causes an uncomfortable state of tensions 

    • Flipside of logical constraint 

    • Psych inconsistency rather than logic; ebleive on some level that 2 things can be true 

  • Belief disconfirmation = being confronted with information that conflict with an attitude that is already held 

    • Eg. Trudeau blackface and claiming to be racially sensitive 

  • Induced compliance = having to justify an attitude that does not hold or persuade ppl to do an unpleasant activity 

  • Free choice = 

• BUT ppl dont like inconsistencies 

  • We want to achieve cognitive balance 

  • ^^ need is strong when there is pressure to resolve such inconsistencies 

  • It varies with the salience of the attitudes involved 

  • Psych predisposition that determine how ok people are with it (some are bothered by it more than others)

• Denial

• Idea (+) ←-------------------> idea (-) 

• We just don’t believe the idea that is making us uncomfortable → rejection 

• OR you become persuaded by the new idea (but generally information you already have is more deeply rooted); updating beliefs 

bolstering

• Have an uncomfortable association 

• Instead of denying the new idea we just add things to outweigh the negative this “yes, well at the same time we have all this other stuff we can think about” 

• Differentiation 

• Element split into 2 parts with dissociation between them 

• Eg. hydrogen bomb testing in valued but poisioning the atmosphere is not → can be differentiated by clean bombs vs dirty bombs 

• Transcedence 

• This negative things exists but i’m going to create a bubble and positively or negative associate with the whole thing 

• Broader organising schema around whatever the thing is 

• EXAMPLE THE LUCKY SHIRT 

• Have both a lucky shirt and lucky tie 

• If you wore at the same time would be twice as lucky but they dont match 

• Thru the lens of cognitive dissonance 

  • Denial - could deny that one of them is lucky and then only wear one (get rid of one) 

  • Bolstering - wear a whole bunch of stuff that matches the shirt while wearing the toes to try to offset the tie not matching 

  • Differentiation - going to decide that the fashion doesnt matter OR could say that you minimize the mismatch and say that you are expressing your own personal style 

  • Transcendence - what is fashion anyways, who cares and it doesnt matter to me 

• How do people choose a strategy? 

• Not usually overt 

• No choice is inherently better - all basically trying to do the same thing 

• Only thing that matter is the individual level of cognitive sophistication (how broad someone’s toolset/capacity to do this is; and how important is this issue to me; level of skill in dealing with abstract, conceptual thinking AND ability to self-criticize) 

• Some require more cognitive energy 

• Denial = least demanding (but not always)

  • followed by bolstering,

    • Bc have to come up with something else to add on to the thing; useful to think of other related things 

  • followed by differentiation 

    • Harder bc need to think in conceptual and abstract terms BEYOND what you are talking about 

    • Basically deconstructing something and splitting into smaller things 

    • And ppl don’t really think conceptually 

  • and then transcendence 

    • Hardest bc you are taking the 2 concepts and building à worldview that subsumes them both and then deciding if you happy or unhappy about it (associate or dissociate) 

• Another factor boils down to the credibility of the new, dissonance causing information (ie WHO is giving you the new information)

  • Profusion 

  • Credibility -- the more different sources are saying things then the more you have to deny it 

  • Plausability -- harder to deny if more plausible 

  • BUT ALL RELATIVE TO THE SOURCES (and strength/rooting) OF THE CURRENTLY HELD BELIEF 

• Transcendence is a rare start in politics bc it requires subsuming a lot of zero-sum choices in ways that are not especially satisfying 

  • Transcendence frames are also not typically provided by political news, which like to emphasize conflict and strife

What is the link between attitudes and behaviours? And what breaks that link and why?

• Disjuncture between what people think and what they say/do 

  • People might consciously deceived others about what their true beliefs are  either thru their actions or opinions

  • People are unaware of their true attitudes 

• The LaPiere Study 

  • Sets out to prove that there is no necessary relationships between attitudes and behaviour 

  • To demonstrate he travels around with an “attractive” chinese couple 

  • He assumes that attitudes towards that chinese are predominantly negative 

    • Chinese exclusion act from 1882 still in force at the time, racial segregation t/o the country 

  • YET they were almost never refused service (1/251 instances) 

  • On the other hand, when he inquired by phone ahead of time the answer was almost always “no” 

    • Happened during teh actual study and also in retrospective questionnaires administered 6 months later 

    • 91% were “no” to accepting a member of the chinese race 

  • What do we make of this finding? 

    • Social Desirability Bias 

      • People will say one thing and then do something else -LP focused on the disparity between opinions expressed over the phone and behaviour revealed upfront at the arrival at whatever establishment 

    • Would it be possible to get this result today? 

      • Laws prevent ethnic segregation in public establishment 

      • But it is extremely unlikely that you could get such a large proportion of people to admit a prejudiced attitude, even over the phone 

      • Even the most bigoted know that their opinions need to be scrubbed before seeing the light of day 

      • So we need to go deeper inside ppls minds 

        • Attitude they know they have and are unwilling to tell us 

        • Attitude people dont even know that they hold

• Roots of teh attitude/opinon/behaviour gap 

• Ppl themselves do not know their true attitude 

• Social norms can change even if people dont 

  • In a world where frank prejudices were allowed there is no reason not to express them to whomever asks 

  • But once that become frowned upon some ppl will feel pressure to conceal those attitudes (SDB) 

  • Emerges around all kinds of issues 

    • Ppl over-report good behaviour like voting and flossing their teeth 

    • And underreport bad behaviour eg, drug use, committing criminal acts, sexual promiscuity etc 

    • Good and bad are not objective BUT they are judgement based on what we think are prevailing community standards

R. Abelson, “Modes of Resolution of Belief Dilemmas”

• MAIN ARGUMENT 

  • Focus on intrapersonal conflict resolution 

  • 2-levels - action level (external responses) and belief level (internal cognitive processes) 

  • individuals encountering cognitive inconsistencies use emotional and motivational constraints—termed "psycho-logic"—rather than pure logic to restore balance and resolve belief dilemmas

 L. Festinger, L., & J.M. Carlsmith. “Cognitive Consequences of Forced

Compliance”

• MAIN ARGUMENT 

  • What happens to a person’s private opinion is forced to do or say something contrary to that opinion?

  • One Study found that the private opino changes to match more the overt opinion but inconclusive 

  • Festinger (1957) proposed that cognitive dissonance magnitude becomes smaller the larger the rewards or punishment is for not publicly stating their belief 

  • ^^ this happens by making private opinion more in line with the public 

  • People who perform a boring task and lie about their enjoyment for minimal rewards experience cognitive dissoncene → they align their opinions with public statements 

  • Small incentives do generate internal pressure

LaPiere, R. “Attitudes versus Actions,”

• MAIN ARGUMENT 

  • Questionnaires not super reliable but they can measure political attitude but do not predict what people will actually do 

T. Devos and M. Banaji, “American = White?”

Reality v. “Mediated” Reality

• MAIN ARGUMENT 

  • Together, these studies provide evidence that to be American is implicitly synonymous with being White

• SDB and evolution of norms 

• Increases of lefthandedness the younger pl are bc we became more ok w it and stopped forcing people to write w the right hand 

• ^ decrease in social pressure 

• Have you read a book in the last year? 

  • 75% yes when asked yes/no and 51% when asked which book they read (if they can't answer counted as a no) 

  • Consistent across groups, not proof of lying but implies it 

• Sherif Autokinetic Effect Experiment 

• Wanted to study SDB - can we demonstrate that people will change their attitude in response to social pressure 

• Adapts the autokinetic effects - the afact that a dot of light protected on a screen in dark room will appear to move

• Tests participants individually on how far the light moved 

• Then tested in groups of 3 with each person having to say aloud how far they thought the light moved 

• Groups tended to converge on a response 

• When in an ambiguous situation a person will look to other for guidance 

• ^^ INFORMATIONAL CONFORMITY 

• Asch Conformity Experiment 

• Examination of social pressure 

• Which of the lines matches the length of the lines on the left? 

  • One correct answer, obvious by design 

• Put a naive participant in room w 4 confederate who had agreed that the anwer was different than the correct one 

• Respondent asked to answer last of the group 

• Found that 

  • 33% conformed w the clearly incorrect answer in majority of trials 

  • Over the 12 trials 75% of respondents conformed at least once and 25% never conformed at all 

  • Most people hve it in them to give into social pressure 

• Very few actually believed the majority opinion was correct - went along to avoid being ridiculed or thought peculiar 

• Speech on campus…some data 

• % of students that would allow six contreversial speakers (diff, categories) 

  • Trasn ppl have mental disorder, BLM is hate group, children should be able to transition w/o parental consent etc 

  • Generally not a majority - and proportion has gone down over time 

  • Social pressure becomes thing that can constrain speech in university environment            

  • Increases in support for taking action against contreversial speakers - high and increasing 

  • Ppl increasingly likely to use social pressure to stop thing they dont want to hear 

• Still… 

  • We should be careful how we frame and think about these findings 

  • On the other hand. . .

    • Only ⅓ conformed and 5% did so consistently 

    • Even those who did explained it rationally through eg. having bad eyesight 

  • Ideally, we'd keep an open mind about what these studies tell us - often taught/emphasized in a sensationalist way

Given SDB how do we get closer to the truth? Survey techniques

• Change how we ask questions 

  • Generally the idea here is distance 

  • People are more willing to admit prejudice by…

    • To someone who looks like them vs someone who doesnt 

    • On the phone (higher levels of prejudice here) vs in person 

    • On the internet vs any live method 

• Change what we ask: instead of asking are you racist? Can ask 

  • Whether they agree that stereotypes w/r/t minority groups are true or false 

  • What most people or people in your neighbourhood think rather than what you think 

  • About things that imply race without stipulating it directly (ie askign about people from china vs asians) 

• Even so remains difficult to ask poeple sensitive questions directly, as a results researchers have turned to 

  • Tricking people: List Experiment 

    • Where respondents are divided at random into direct and veiled categories - nobody knows this 

    • Enumerator reads statements and respondents have to say how many made them upset 

    • Second groups has all same things as group 1 BUT includes race question 

    • If this sensitive item didnt matter then on average the results would be the same across groups 

    • Inference is that the race thing is something that people would not tell them SO there is some element of prejudice present 

    • Can estimate the proportion of people for whom the race question mattered 

    • Can only infer about groups 

  • Subtle field experiment: Audit Studies 

    • Researchers send out of cvs to apply for real world job ads 

    • They vary the name of teh CV so that the reader will be cued to the fact that the applicant is white or an ethnic minority 

    • Assuming all else on the cv is the same if callbacks are lower for the thnic names, this is evidence of racial bias 

    • What makes this work is that employers only see ONE cv 

    • Eg. 2004 study contrasted Greg and Emily with those for Jamal and Lakisha 

    • Callback for white names was 9.65% and for african american names were 6.45% and similar gender difference 

  • Subtle field experiments: Social Media 

    • Recent paper assessed subtle racial bias in acamdeic networks using twitter 

    • Ide to generate fake grad students, experimentally vary their characteristics (gender, race, institution) and then ahve them follow a bunch of economists on twitter and see who gets followed back - how does this vary based on different characteristics 

    • Results 

      • Females got followed the most and not huge differences overall 

      • Black amle-lower ranked was followed the least; top ranked schools tend to be followed more 

      • Biggest differences are prestige of institution and gender 

  • Measuring things other than what people say: The implicit association test 

    • Asks you to pair two concepts (eg. young and good, or elderly and good) 

    • The more closely associated teh two concepts are the easier it is to response to themes as à single unit 

    • So if young and good are strongly associated it should be easier to respons faster when you are asked to the same response to these two

    • If say, elderly and good are not strongly associated it should be harder to respond fast when they are paired 

    • This gives measure of how strongly associated the concepts are the more associated the faster you should respond 

    • Interpretation is described as automatic preference for X based on response speed

• Project Implicit 

• Critiquing the IAT 

  • Aren't stable at the individual level the way you'd expect a lurking trait to be 

    • Things are only of concern when they are stable traits or characteristics 

    • High consistency within tests BUT low consistency across them (test-retest reliability) 

  • Questionable relationship to real-world behaviours 

    • Not no correlation but fairly low correlations between implicit attitudes and explicit behaviours and judgements 

    • No added power to predict behaviour over and above more explicit measures (aka incremental validity) 

  • No evidence that changing people's level of implicit bias has any impact on explicit behaviours and judgments 

  • Possibility that IAT scores are not measuring implicit bias, so what is it? 

    • General memory associations, cognitive flexibility/inflexability 

    • Contextual activation or familiarity with stereotypes rather than a stable internal prejudice 

    • Personal cultural knowledge rathn tahn hidden personal prejudice 

    • Statistical noise

• Anonymity is not always anonymity and using big data to reveal illicit attitudes

• All google searches are anonym ous as far as to anyone who is not google so people search information they would eb weary about telling another human being 

• As a result aggregated search data cab shed valuable light on actual community standards 

• Racial Animus 

  • Aggregation of searches that include the n-word reveals interesting picture if american’s attitudes -- where is the prevalence of certain search terms is 

  • South is about what you would expect but NE appears no less racist on this measure 

  • And plenty of activity happening in the west (supposedly liberal) 

  • How is this politically relevant? 

    • As state becomes more liekly to search in racially charged terms support for Obama goes down 

    • Politically relevant bc support for Black candidate goes down but if you ask people that they would say they are comfortable voting for them 

• Search for T, voting for T 

  • Search rate for term testosterone related to trump support --- positive, upward slope 

  • Search rate for feminism related to trump support -- negative, downward slope 

• Google searches over time 

  • Frequency of searches for homophobic, racist, sexist jokes gone down overtime up until 2017 

  • Could be being exposed to it on algos, social media etc 

  • Decline overtime is significant and not obviously tied to evolution of norms around social desirability 

  • Decline is probs understated by these measures of you think about how the population of internet users has evolved

• The Bradley Effect 

• Frank prejudice can be politically relevant if it translates to political choices 

• SDB can be pernicious in a different way 

• Bradley effect 

  • Name after LA mayor tom bradley an African-America who lost the 1982 governors race depite being ahead in the voter polls 

    • Respondents who said yes could have said so bc saying no would amke them appear racist 

    • However once confined to the secrecy of polling booth some vote against him 

  • Ppl worried about Obama for the same reasons ^^^ 

  • If polling departs then we can see that there is bias of this kind 

  • More social desirability among democrats

• 
  

S. Iyengar, M. Peters, and D. Kinder, “Experimental

Demonstrations of the Not-So-Minimal Consequences of

Television News Programs.” (JS)

• Experiments sustain the suspicion that media provide compelling descriptions of the world ppl cannot directly experience and this affects which problems viewers see as important 

• Eg. problems featured in the evening news are weighted greater in evaluations of presidential performance 

• Results 

  • Participants rated importance of problems 

  • Participants who were exposed to exp #1 exhibited increased concern about US defense of the next daays (ranked 6/8 concerned then jumped to 2/8) 

  • Strongly supports the classical agenda setting hypothesis 

  • Exp #2 they rated the performance of the president 

  • we then correlated judgments of President Carter's performance on a particular problem with rating of his overall 

  • ^^ came back the media presentation doe alter the standard ppl use in evaluating the president 

  • More new coverage → more information → ppl conclude the problem is important 

  • Lipmann’s suggestion is sustained but uncertain how long those effects persist 

  • Political consequences of the media’s priorities are ENORMOUS - candidate my ignore problems that arent getting much public traction and candidates taht are not taken seriously by the news may not be taken seriously by anyone else (advantages and disadvantages depending on who you are)

• 
  

- Krosnick, J., & Kinder, D. “Altering the Foundations of Support

for the President Through Priming.” (JS)

• The more the media pay attention to a thing the more people aare primed with it and the more  they incorporate what they know about that thing into their overall judgement of the president 

• RESULTS 

  • Public opinion on foreign affairs did become more imprtant for assessments of prez performance in the immediate aftermath BUT after finding out foreign affari b important in evaluating and importance of onctras increased substantially 

  • Theory of priming stands up well to examination 

  • ory stands up well to this examination. Citizens questioned after the revelations held President Reagan to an altered set of standards, and these alterations can be directly traced to the changes in the media's agenda

  • ^^ aspects of public opino more directly implicated by news coverage 

  • More apparent among novices than experts w/more focus on performance than character 

  • Changes are dominated by the least informed

- Mutz, D.C. “Effects of ‘In-Your-Face’ Television Discourse on

Perceptions of a Legitimate Opposition.”

• Asking how american acquires the impression that their politicla foes have some bias for their views and are legit opposition 

• Mass media/tv menas of being exposed to opposing perspectives 

• Experiments suggest that TV has capacity to encourage greater waaremenss of opposition BUT characteristic of televised political discourse causes audiendes to view oppositional perspectives as less legit 

• RESULTS 

  • Uncivil exchanges = more arousing 

  • Those who watched the program consistently knew more than those who did not 

  • . When the uncivil candidates were shown in close-ups, the gap between evaluations of the two candidates was exacerbated, just as close-up disagreement from another live human being intensified attitudes in the laboratory. However, the very same means for subjects watching the exact same event from a medium camera perspective did not differ.

  • Relative to the control group, the people who watched televised political discourse came away from the experience better able to recount the other side’s arguments as well as their own

  • Close-up perspectives on uncivil exchanges of political views are arousing for viewers to watch, and thus help to draw audiences to these programs, as well as to encourage attention and retention while viewin

  • ome at the cost of lowering regard for the other side. The “in-your-face” intimacy of uncivil political discourse on television discourages the kind of mutual respect that might sustain perceptions of a legitimate opposition. Here the pattern of findings is quite consistent; close-up perspectives on uncivil discourse routinely damage perceptions of the candidates and issue arguments that subjects are already prone to dislike; that is, attitudes toward the least-liked candidate, and the perceived legitimacy of rationales for opposing issue positions. 

  • ^^ same pattern does not occur for the preferred candidate

 Prior, M. Post-Broadcast Democracy: How Media Choice Increases

Inequality in Political Involvement and Polarizes Elections, Chapter 1.

• attention to mass media required to explain political behaviour

• Systemic treatement of how the media environment affects political behaviour 

• 
  

• PPL DO NOT VOLUNTARLY CONSUME LOTS OF NEWS 

  • As far as Downs’s first path is concerned, people who do not intend to learn about politics still do so because they cannot find exactly the media content they prefer.

  • Floating voter → voters whose decisions are not noticeably influenced by a stable liberal or conservative ideology or a strong adherence to one party. As a result, they seldom approach an election with a firm sense of whom to vote for and do not always vote for the same party. 

    • These ppl are less educated compared to those who tend to vote for the same party 

  • The amount of political info floating voters encounter affects their tendency to vote at all (high stimulus vs low stimulus elections) 

  • 
    

  • SOME PPL VOTE OF THEU ENVIRONMENT REMIND THEM OF THE ELECTION AND PROVDES THEM WITH SIMPLE REASONS TO PIC ONE SIDE OR THE OTHER 

Black female candidates in gubernatorial election

• Looks at bias against both black and female candidates in gubernatorial elections over time 

• Found that 

  • There is a Bradley effect of Black candidates in the 90s (did better @ the polls than votes) BUT the effect disappeared by 2005 

  • ^ can argue norm revoluation - behaviour caught up the rhetoric

  • Found that there was not effect for female candidates 

• So what does it all mean 

• Attitudes and behvaiours can be meaningful different 

• Much of this stems from the pressure ppl feel to conform 

• Somer things to considers 

  • How are we measuring implicit bias nad what does it really mean? 

    • IAT and milliseconds 

    • Experiments and the shifts to comparing across groups rather than measuring bias at the individual level  

  • Double-edged sword: mostly discussed ppl censoring themselves in a “bad” way bc of SDB - is this worth talking about 

  • Why do care about attitudes if people modify their opinions/behaviours to be more socially acceptable? 

    • BUT biases manifest themselves in subtle ways, and if people are not aware of these biases, they cannot actively censor themselves 

  • If implicit and explicit bias are correlated , its unclear exactly what added explanatory mpower measuring the former gives us over the latter 

    • The differences is rarely pivotal to aggregate outcomes 

    • The differences doesnt give us much added leverage to explain outcomes across cases (consider google search example)

• Mediated reality 

• Most of the time we come into contact with the political world directly 

  • W industruailization world has gotten bigger 

  • About ability to actually experience the world via direct access tot he physical senses has not 

  • So need to understand mediated reality on attitudes, opinions, and behaviours 

    • Lippmann -- relationship between what others have reported and what we can imagine 

    • Is it the cases of the medium of the message? 

• Assessing Risks 

  • Causes of death in us laregly cancer and heart disease, <0.1% of deaths are from terrorism, but the causes of death the NYT reports on are 35% terrorism releated 

  • Only 3% support partisan violences but 63% of Americans say that political rhetoric is encouraging violence a lot

• Psychology of the crowd 

• Many social thinkers understood industrialization/modernationzat as a de-stabilizing force for amss society bc it erode traidiotnal values and took on new ones 

• How to understand the psychology of an untethered crowd? What is guiding their behaviour? 

  • Gustave Lebbon(1896)  contended that ppl in crowds feel free to gratify savage and destructive instincts 

  • Freud extende this to empahsize the role fo group leader; bonds between crowd members depending on common allegiance to him 

  • Krohuaers - modernization leads to breakdown of primary group ties --- allows ppl to move from one idea to another 

• Concerns about susceptibility of the rootless crowd suggested that they might be easily manipulated by demagogues -- no limit to absurdities 

• The world is bigger than their sense and institutions that made sense of things are no longer operating 

• Hang Mike Pence? 

  •  Were most people really thinking they were going to murder the VP? 

  • No, most ppl were not 

  • BUT crowd psychology can overpower individual rationality

the massive effects model

• MEM= with the popularity of radio and TV, there was intense fear that the public would be turned into a pack of mindless drones, easily controlled and manipulated (tech will permit demagoguery) 

  • Unlike previous newspapers which at least had to be read the concern existed that electronic media could essentially hypnotize ppl into action 

• Eg, Halloween 1938 

  • WAR OF world broadcast convinced 1000s of ppl that the planet was being invaded by aliens 

  • Panic was brought by one single broadcast w no evidence of emergency from outside the broadcast 

• Around the same time, widespread concerns about the potential effects of political propaganda (eg. Germany’s use of modern media to entrance the German people) 

• Shows that how seriously ppl took the power of demagoguery and the ability to mass media to tansi the message to larger sphere 

• Institutes of Propganda Analysis 

  • Designed to alert the world about the tricks of the trade 

  • Eg. speeches entrance you and you cant talk back or so anything about it 

  • Father Coughlin and his antisemtimic remarks and patriotic rallying cries 

    • Name calling, testimonials (attributing an idea to a good or evil source), bandwagon (saying that everyone believes something)

The minimal effects model

• All the fears about mass vulnerability to propaganda and media effects more generally relied on several key assiumptions 

  • Political propgande highly persuasive 

  • Propagandas success depends on how effectively the trick os the trade are marchslled 

  • Most importantly:the audience is captive and gullible 

• By the 40s and 50s had generally found that exposure to the news media did not really change ppls minds about anything at all 

  • The first public opinion studies on vote choice placed media effects in a central role 

  • Media has little persuasive power because audiences are largely self-selected 

    • Those paying atteiont tended to alerayd pay attention and itnerpreted selectively - just reinforce what you already thought 

  • Skepticism reinforced by social psychologists studying the effects of US gov attempts to indoctrinate new draftees with orientation films during WWII 

    • Studies found that films were effective and getting across information about whats going on, battle lines etc 

    • But did not produce change in behavioural intentions (ie wanting combat or non-combat roles) 

    • All this despite the fact that the audience was captive rather than tuing in an out as they would with political campaigns 

• MINEM 

  • Media propaganda rarely changes ones attitudes but instead serves to strengthen existin attitudes

• Why doesnt the media persuade ppl? 

• People have to received AND accept the message 

• Receptions 

  • Most ppl dont seek political news 

  • Even when its one they dont necessarily pay attention - how much do they really remember? 

• Acceptance (assuming dissonance)

  • Once incoming info is received it is evaluated base don what prior attitudes were 

  • Consistent infmroaptoin is readily accepted and depending on level of commitment to prior attitude discrepant info is rejected 

• In either case existing attitudes tends to be strengthened 

• Exception that proves the rule - rally around the flag effect 

• When WHO declares a pandemic public support for leadership in all kinds of countries get more popular (bc people turn to them when they feel threatened)

• Trump did not get more popular nor did Bolsonaro 

• In a sense this is media persuading ppl that their leaders are doing a good job 

• People paying more attention to political information 

• Obstacles to reception and acceptance went down 

• Heighenedwith 

  • How severe the issue is 

  • Breadth of impact on society 

  • Degree to which the media consensually supports the leader 

• But this is relatively rare 

• Not-so-minimal effects: Agenda setting 

• Media outlets invariably highlight some issues while downplaying or ignoring others; complex world limits ability to learn about it 

• Agenda setting = process of choosing what and what not to cover 

• Not persuasion per se and involves focusing the public’s perception on a particular issues at the expense of the other w/o makes deliberate effort of telling what tothink BUT what to think about 

  • Lyengar and kinder shower the making ppl watch coverage of certain issues rate those as more important 

• Reasons to be cautious 

  • Out in the world ppl freely choose their media and media act to maximize viewing/reading audience -- ideas of spurious correlation and reciprocal causation 

  • Agenda-setting effects can occur outside of or independently of media coverage 

• Eg Wash Post vs Fox NEws 

  • Very different headlines ie agenda setting 

• Examples suggest different editorial emphasises which just suggests that they tried to set the agenda, dont know how seriously people actually took it SO need to survey diff viewers btu evidence not deifnitives

• Priming 

• News may influence the relative weight os standards by which governments, presidents, policies, and candidates for public office are judged 

• Ie evaluating how things are going 

• ^^^ priming; occurs when media changes the evaluative weight of some considerations over others 

• Eg imagine someone’s Trudeau’s judgment depends on his facial hair and the economy

  • Pre-covid facial hair tipped the scale and post-covid the economy weighed more 

  • Asking the same question but things are being weighed differently (ie evaluative weight changes) 

  • 
    

• Doesn't necessarily push people one way or another (that would be persuasion) 

• EXAMPLE 

  • Was racism primed by the Donald Trump candidacy/presidency? 

    • Did trump prime people to make decisions more on racial lines 

  • Panel study, same respondents re-interviewed at numerous points starting in 2012

  • Asking in first survey → questions tied to anti-black prejudice and score them based on explicit negative racial attitudes 

  • X axis is prejudice (increases left to right) 

    • Graphs howing relationships between how prejudiced the person is how they favour an african american candidate 

    • 2012 -- kind of flat, not big difference ib voting between prejudiced and non-prejudice BUT that slope gets steeper overtime leading up to the 2020 election 

    • Can infer that the influence of anti-Black prejudice has become stronger -- racist idea primed by trump that become rpedictive of who they are going to vote for 

  • Then starts asking how are they going to vote 

Framing

• The way the public makes sense of the world/these alternative/schema 

• Message frames issues by encouraging readers/listeners to emphasize certain considerations above other when evaluating that issues, depending on the priority to various considerations 

• Occurs when individuals arrive at different positions on the issues, depending on the priority that is given to various considerations 

• Framing about thinking about the story 

• Eg.

  •  newspaper editorial defending a hate rally  of free speech rights may move readers to facour the rally by heighting the evaluative weight of speech concerns 

  • Editorial about public safety concerns may lead readers to give priority to maianing social order and turn them against the rally 

  • Even if they use the same case facts 

• EXAMPLE 

  • Espteins birthday book 

  • 2 framings 

    • “Epstein birthday letter with trump’s signature revealed” 

    • “Clinton’s message to Epsten revealed as GOP slams dems for ‘cherry picking’ “ 

  • Immigration 

    • Can we reframe the way that people think about legal immigration? 

    • Respondents asked if they should let more or less immigrants in either referring to skilled immigrants, family members, refugees 

    • People more likely to accept immigrants based on these 3 conditions compared to 50% of the control group says they want to decrease -- goes down to ~25% for each group -- ppl becomes much more moderate 

  • EXAM QUESTION: do these ppl seem to know what the difference is and how do they apply it?

framing vs priming

• Harder to see in practice than in theory 

• Agenda setting provides new information and priming does not → always knew that X was important but we dont always connect it or think about it 

• But is it possible to prime w/o providing new information? 

• The distinction between agenda setting and framing is in framing there is no new info presented -- instead offers perspective to organization information that ppl already have 

• Framing vs priming 

  • Mostly emphasis

  • Studying priming generally assume that evaluation will be made in response to whatever criterion is primed 

  • Those who study framing are more interested in the effect of competition between ompetting frames or ways of thinking about information

• Ideological Polarization and Media Use 

• Who watches what and why that might matter 

• Peoples ideolgoy seems to predict what kind of media they consume 

• Eg left -- PBS, BBC, right -- Fox etc 

• Liberals are more heterodox in choice of media, more trusting of a variety of sources -- tend to name a variety of sources when asked, trusted 28/36 news orgs

  • But conservatives more likely to be homogenous and trusted 24/36 news orgs SO concentrate more in one area, distrust 20/30 sources 

• Liberal Bias in media 

  • Nigel Frage saying taht hte BBC is infected with left-wing bias 

  • But he has appeared on BBC question time 38 times 

  • Some studies show taht the media is quite liberal (at least the people in ti identify themselves largely as democrats (32%) vs republicans (3%) 

• But is the media causing anything? 

  • We know viewer ideology is correlated with what news source ppl choose 

  • We also know that the media has become more polarized in terms of content - fox got much more conservative over the years, CNN more centre, and MSNBC further left 

  • One interpretation of this is that media exposure is altering people attitudes and behaviours 

  • But another possibility is that people simply choose whatever source of news lines up with their predispositions 

    • Ppl are free to watch/read what they want 

    • Media companies adapt their ideological message to their target market 

  • Referred to as selective exposure 

    • Points to an issue that most lab experiments on media exposure failed to address until very recently 

• Compounding effect of social media: prevalence 

  • Majority of Americans get news from social media, higher than even before

  • Increases in tiktok, facebook, youtube etc 

• Partisan slant 

  • Most polarsized in instragram but the lean is democracy 

  • Most social media users still tend towards democrat leaning on facebook 

  • But twitter now largely repblicuan - posting largely republican

Algorithmic politics

• Legacy media has become more ideological to a degree 

• Seen a profusion of websites populating the extremes of both sides of the spectrum 

• Social media compounds the problem by algorithmically concentrating what ppl see based on what the platform thinks they will click on - business of the algo 

• Computation analysis of elections up until 2024 

  • After musk bought twitter republican content skyrockets on twitter, democrat content quite low comparatively 

• Researchers looked at the effect of algo by comparing attitude change scross two twitter groups assigned at random 

  • Viewers of standard following feed 

  • Viewers of for you feed 

  • Difference emerged after 6-7 weeks for those switched in both directions 

  • Found that 

    • Ppl who went from follower to for you feed → moving more republican based on questions asked about policy, priorities etc 

  • Worrisome because the reverse doesn't really apply, the algo matters and changes views on things 

• Are you making things worse? 

  • 2015 study of 10M facebook users 

  • On one hand ppl have diverse networks and this might expose them to alternative viewpoints 

  • Divided by libs, cons, mods and how amny of their friends are on the other side 

  • Ppl mostly had friends similar to them but a good amount of ppl on the other sides 

  • BUT ppl are less likely to share or click on new things they encounter from ppl/sources of an opposing ideology 

  • Conservatives are more likely take cross-cutting content and see it, click on and read it 

  • Would appear that libs are less tolerant of cross-cutting information than conservatives 

  • Suggests that there is certain breed of conservatives that are hate posting or could be open to more diverse information . . .

    • OR there could just be more liberal information out there so ppl more likely to encounter it

Media negativity: If it bleeds, we read?

• Pronounced negativity bias both in what is covered and what we read/share 

• Ppl are much more likely to click on and read negative articles - more negative words, the more likely the click and share becomes 

• Increase in positive words less likely to reshare and read 

• Ppl more likely to read and share negative news

• What if we disconnect people? 

• Users randomly deactivate from FB and IG in runup to 2020 election 

• Nothing much changes 

• Turnout doesnt changes but ppl who got off FB were 3pp less likely to vote for trump 

• IG no statistically significant effects 

• BUT there was big decrease in political participation - due do lack of political postings SO not demobilizing

Information v misinformation

• Misinformation is everywhere in the vernacular and experts see it as pre-eminient threat to democracy 

• Diff kinds of misinformation identified as concerning in surveys 

  • USA: politicians spreading it, SM failing to stop it, AI being used to deceive ppl etc 

• BUT really hard to conceptualize what misinformation it 

• The notion of untruths is intutives but poiticail discourse lies in the middle 

• AND ppl dismiss things that they dont like as being misinformation - convenient fig-leaf for it 

• Talarico 

  • Gets sick on the campaign trail and posts claim that he was at a mental facility 

  • Can agree that this is probs misinformation

What is MI exactly?

• Factually false, intention to mislead, 

• BUT 

  • Ffacutal falsity has an element of subjectivity, perspective, and framing in most of politics (eg. discourse on violent crime in democrat run cities) - not necessarily untrue bc there are crimes in most cities 

  • FF often gets tied up with nrmal stuff that cant be disproven (eg. windmills are bad or windmills kill birds) 

  • Easy to conflate sources an outlet: is a mainstream outlet circulating fake news when it accurately reports on something outrageous trump said in his speech? - is that misinformation or not? 

  • On intention to mislead outside of obvious scam email how do we know what anyone’s intention was in circulating something? 

• How true and false info spreads 

  • Study looked at twitter retweets 

  • False rumours went more viral that true ones 

  • False rumours are also more connected to negativity 

• Spreading misinfomraiont abe being sanctioned for sperading misinformation 

  • Biden hashtags vs trump hashtags 

  • More sharing of low quality false stuff on the right and volume of this kind of information is much higher

• Does correcting misinformation work?

• Look ed athow ppl response to fact checking 

• Generally fact-checks do reduce beliefs in misinformation and this can be long lasting - very strong correction effect 

• The message still has to be received by the hearer of the information -- are the people that need to be corrected listening/can we get it to them? 

• We are fact-checking but who is listening? 

  • Information diet is largely self-selected - reduces the likelihood of the misinformed actually getting the correction 

  • ^ same issue as in the context of other media effects 

• Who is the fact-checker, and on what basis will it be viewed as legit? 

• Other predispositional factors 

  • Political affiliation and other political identities alter the likelihood that the correction will be accepted 

    • Conservatives less trusting of mainstream media and more likely to call it fake news 

    • Conservatives tend to be less able to recognize fake news than liberals 

  • Cognitive ability 

    • Ppl scoring lower on cognitive tests tend to be credulous w/ fake news and less receptive to corrections 

    • Heavy social media use increases the likelihood of individuals endorsing politically-motivated fake news 

• Not everything can be fact-checked on a conceptual eleven even if doing so might work on people; most of politics is not a battle over the truth value of objective facts

RAS model

• RAS model imiagianes the process by which à polticial message/consdiereation is absorde into à person’s mind and by which it ultimately become part of an epxressed opinion 

• R 

  • Reception -- likelihood of person receiving a political message which in turn depend on various psych predispositions (eg. interest)  as well as socio-econ (eg. tech access) and social factors (eg, friendships) - basically anything that could alote the probability of someone being exposed to it 

  • Q: Consider an adept viewer of MS now, and Fox runs a story about this will it change their mind about X? -- first thing you say is ok are they even giong to get the message

• A 

  • Acceptance - likelihood that the person will accept the message, dependon whether is is consonant or dissonant w respect to what the person already believed 

• S

  • Sampling - liklihood that the message, once accepted, will be sailet (at the top of their head) at the moment somebody askd them, depends on the extent and nature of media coverage (conditions reception ofc) and subsumes priming and framing effects 

  • Eg. if you are v strongly partisan then that is going to influence your sampling of certain ideas when asked the same question as another affiliation

• Rational choice theory + how a rational choice is made

• Individual level model of decision making (no such thing as group rationality) 

  • What is best for the groups is not necessarily best of each person 

• 3 elements 

  • Feasible set of all course of actions which are rationally believed to satisfy various logical, physical, or economic constraints 

  • a rational set of beliefs about the causal structure of beliefs about what choices will lead to what outcomes and why 

  • a subjective ranking of feasible alternatives in terms of utility associated with each outcome 

    • Utility must be expressed as some kind of common or comparable unit 

    • Ranking must be consistent, in that preferences will not bounce around haphazardly 

    • Must be transititve À >B and B>C then À>C

  • The rational choice is the one that maximizes the utility fo the individual making that choice 

• How a rational choice is made 

  • Basically cost-ebenfit caluacaiton - compares costs and benefits of various options and pick the noe thats highest 

  • RC  says nothing about which choice someone will make whitney alternatives have equal expected utility 

  • RISK AND UNCERTAINTY (makes things complicated)

    • Can assign numerical probability to a given outcome 

    • In the case of uncertainty, we cannot

    • In these cases RC offer very littel leverage

• Maximizing utility under risk 

• Risk is traceable mathematically all be do is multiply the utility associated with each alternatiave by the probability with which it will occur 

• Expected utility: E(U1) - P(B1) -(C1)

• EG 

  • a game -- roll a fair side die, and if you get 1-5 you get $20 but if you roll a 6 you lose $100) 

  • MATH:  E(U1) - (5.6)(20) -(1/6)(-1000)

    • EU= -150 

• EG Proximity Voting Model 

  • Assumed that ppl will choose candidates based on whichever is closer to them along the liberal-conservative ideological spectrum 

  • The decisions is parametric and based on a single estimate of ideologically proximity 

  • But form the standpoint of parties, there is a strategic question of where to place thesmseves to get as many voters as possible 

  • SO liberals may move themselves more conservative under the assumption that voters are going to vote ideologically SO everything further left of their position will vote for them btu will not go past the repubs 

  • ^ but the repubs also have the same incentive 

  • As a result both of them move to the centre = MEDIAN VOTER THEOREM 

  • Assumes no cost to the parties if they change their position

rational choice theory and voting

• EG Prisoner’s Dilemma 

  • 
    

  • Dominant strat is to betray - each player's best option is to betray 

    • Best thing is to betray to maxmize one’s own utility bc if you betray and they betray you you serve less and if you dont and they do .. 

• EG Paradox of Voting 

  • Assume 

    • People have some ideological/polciy ideal point 

    • People will only vote if benefit of choosing one cnadiates over the other exceeds the cost

    •  voting decision = B-C, will vote if >0 

    • But you also have to consider whether your vote is actually influential. Why? 

    • If not influential, the incentive is to free ride - most elections the impact is negligible 

    • On these assumptions nobody should vote 

  • BUT people still do even though its not rational and there are other things that would give you more utility 

    • Factors that lower the cost of participation? - many countries you get registered automatically, some places it is mandatory 

    • There are elements of the psychology of voting that arent rational but there also exists things like social pressure 

  • Voting decision = P(B) -C+D+E

Rational choices are difficult because…

• Decisions are complex, we are generally working with imperfect information; decisions are time-sensitive

• Most political deicions we make are relatively low salience 

• Heuristics 

  • May lead to biased decisions based on selective perception 

  • BUT there are fundamental heuristics -- that are systemically part of human nature 

• Ppl have very little knowledge of the sociodemographic world around them 

  • Sometimes this reflects broader information environment (ie what is visible to you may lead to over/under estimations)

  • sometimes its just the way people pick numbers when they really dont know something 

• Ppl are fundamenetlal hampered by cognitive bias 

  • Thought experiment 

    • Imagine you are on a game show and presnseted w 3 closed doors, behind one is a car you want and the other two have goats 

    • You choose one door but it remains closed, now the host who know what is behind each door as if you want to switch 

    • The correct answer is switching bc it yields a ⅔ change of getting the car whereas staying gives you only ⅓ 

    • See slide to visualize 

  • Why we consistently overrated the change of the door they already picked? 

    • Endowment effect - in which ppl tend to overvalue the winning probability of the door already choses - already owner 

    • Status quo bias - ppl prefer to make the hocie they already have 

    • Errors of omission vs errors of commission - all toher things being equal ppl prefer to make errors by inaction (stay) as opposed to action (stay)

prospect theory and approach to risk

• Risk aversion 

  • We would rather take $100 guaranteed than a 50% shot at $200 (and 50% $0) 

• In the realm of losses we have implied risk acceptance 

  • We would rather take a 50% shot at alosing nothing (vs 50% $200) than a guaranteed loss of $100 

• If you feel like you have something to lose you are much less likely to take the risk 

• Expected utility doesnt matter but the domain of winning or losing does matter a lot