MacKuen and Brown - Political Context and Attitude Change

Using ANES data, authors demonstrate that environment shapes the way citizne views politics. County-level and residential neighborhood level contexts. Citizens' evaluations of candidates and parties are most directly influenced by what their neighbors are saying at that moment. Citizens' self-identification evinces sensitivity to the more stable partisan character of the environment. Socail setting is an intervening mechanism in the communication system and is ont merely an exogenous source for political information.

Is it sufficient to think of the citizen as an individual acting in a collective environmen, or must we taking into account the citizen also being defined by the environent?

Background
Interest in contextual effects stems their serving as evidence for interpersonal influence. The mechanisms of social influence have been variously portrayed.

Little is known about how mutual social influence operates. It could be that individuals adjust their views in accord with political norms in the environment, or that interpersonal influence does nt depend on long-term conditions but instead on hot communication - those communications that are current ongoing. Hot communation could be about the tendency of members in the environment to offer views consistent with their own beliefs, or an individual could take information from externl sources (media), form a personal understanding, and then place it before peers for confirmation. In the latter case, context is an intervening agent.

The Data and Analytic Strategy

Looking at ANES wave data.



Looks at macrocontext of the R’s county. Microcontext is R’s slef-report of perceptions of specific neighbors’ partisan loyalties and vote intentions.

Change determines causality. If the social environment affects attitude formation, then there should be an empirical trace in sequential observations of individual views.

You would expect that individual will align their views with those dominant in their political milieu.

The Macroenvironment: The County
The county is not a great indicator, but maybe the best available. Two questions about the macroenironment: (1) what are the magnitudes of the contextual effects occurring over time? (2) During which period of the campaign year are the contextual effects more pronounced?

The Magnitude of Contextual Effects

Republican context has the best explanatory power of many variables, exepctions of initial feeling and R’s partisan identification. When you looked at time change, the effect is more pronounced.

---this doesn’t show the mechanism of social influene, but it indicates that context on attitude change is statistically evident and substantively important.

Timing of Context

At what stage of the electoral calendar does context matter most? The early part or in the heat of the late campaign?

Timing is more difficult to figure out. “Context is complex.”  It varies across people and parties – partisans will respond to information differently.

Microenvironment: In the Neighborhood
Authors admit that neighborhoods really aren’t that great a proxy. But there is some substantial minority that can be studied. If we can identify a social effect here, we have evidence for social effects at work or with family and friends. We look for social communication phenomena where they lie in our data and guess that they exist elsewhere.

<p class="MsoNormal">Identifying Social Influence

<p class="MsoNormal">When citizens discuss politics, perhaps at the most casual level, the social environment shapes the substantive content. Our expectation is that citizens, taking cues from their conversational partners, will encounter and more readily accept information consistent with the political character of the environment.

<p class="MsoNormal">The mathematical modeling for conversations among different partisan groups of neighbors affecting candidate evaluation is not as crisp as you expect. But, the data correlating neighbors’ vote intention and change in R’s evaluation of candidates is quite strong.

<p class="MsoNormal">While the contextual effect of the macroenvironent (the couty’s partisan coloration) is generated by some historical tradition or the media, but the neighbor/friend contextual  change can be attritubed to social interaction.

<p class="MsoNormal">On Micromechanisms

<p class="MsoNormal">What is more important? The partisanship of your neighbors – a general background feature – or their intention to vote – a form of hot communication? Clearly, when it comes to shaping information about the candidates or the parties, hot communication is much more powerful. This is more consistent with the social communication rather than the social cues model of social influence.

<p class="MsoNormal">Long term political history of the county is at best a measure of local partisan patterns – and in light of this research, this may not be so relevant, as it is not the fundamental character of one’s social milieu that matters; it is what people think and say at the moment that carries the day. It would seem that your understand of your neighbors’ voting intention is about as powerful as your self-proclaimed partisanship.

<p class="MsoNormal">The power of the neighbor vote intention data suggests w/r/t hot communication that the social environment is an intervening force that receives and processes outside information – Dems are not getting much bad news about Reagan confirmed by friends, because there wasn’t much bad news about Reagan. The “social reality-test” model predicts this asymmetry – the idea that news is presented for friends for confirmation or disconfirmation.

<p class="MsoNormal">Context and the Development of Individual Partisanship

<p class="MsoNormal">Partisanship is strong and durable identity characteristic – but it tends to move towards friends’ partisan orientations or toward the content of their conversations. Authors admit that neighborhoods really aren’t that great a proxy. But there is some substantial minority that can be studied. If we can identify a social effect here, we have evidence for social effects at work or with family and friends. We look for social communication phenomena where they lie in our data and guess that they exist elsewhere.

<p class="MsoNormal">

<p class="MsoNormal">Identifying Social Influence

<p class="MsoNormal">When citizens discuss politics, perhaps at the most casual level, the social environment shapes the substantive content. Our expectation is that citizens, taking cues from their conversational partners, will encounter and more readily accept information consistent with the political character of the environment.

<p class="MsoNormal">The mathematical modeling for conversations among different partisan groups of neighbors affecting candidate evaluation is not as crisp as you expect. But, the data correlating neighbors’ vote intention and change in R’s evaluation of candidates is quite strong.

<p class="MsoNormal">While the contextual effect of the macroenvironent (the couty’s partisan coloration) is generated by some historical tradition or the media, but the neighbor/friend contextual  change can be attritubed to social interaction.

<p class="MsoNormal">On Micromechanisms

<p class="MsoNormal">What is more important? The partisanship of your neighbors – a general background feature – or their intention to vote – a form of hot communication? Clearly, when it comes to shaping information about the candidates or the parties, hot communication is much more powerful. This is more consistent with the social communication rather than the social cues model of social influence.

<p class="MsoNormal">Long term political history of the county is at best a measure of local partisan patterns – and in light of this research, this may not be so relevant, as it is not the fundamental character of one’s social milieu that matters; it is what people think and say at the moment that carries the day. It would seem that your understand of your neighbors’ voting intention is about as powerful as your self-proclaimed partisanship.

<p class="MsoNormal">The power of the neighbor vote intention data suggests w/r/t hot communication that the social environment is an intervening force that receives and processes outside information – Dems are not getting much bad news about Reagan confirmed by friends, because there wasn’t much bad news about Reagan. The “social reality-test” model predicts this asymmetry – the idea that news is presented for friends for confirmation or disconfirmation.

<p class="MsoNormal">Context and the Development of Individual Partisanship

<p class="MsoNormal">Partisanship is strong and durable identity characteristic – but it tends to move towards friends’ partisan orientations or toward the content of their conversations.

<p class="MsoNormal">Self-identification relies on a different sort of social input than candidate evaluation. Self-id depends on imitative rather than cognitive learning.

<p class="MsoNormal">

<p class="MsoNormal">Self-identification relies on a different sort of social input than candidate evaluation. Self-id depends on imitative rather than cognitive learning.

Conclusion
<p class="MsoNormal">Context operates through interactions with identifiable friends and neighbors: concrete personal relations rather than amorphous community norms are the proximate cause. The content of contemporary conversations, rather than stable characteristics of the social milieu, generates the contextual phenomenon.

<p class="MsoNormal">The dynamics of context may be complex. It could be potentially explosive, but the more stable political character of the environment dominates long-term change.

<p class="MsoNormal">The overall picture of social influence is interdependence. The effectiveness of context depends on the sorts of information generated by the external politics of the moment.

<p class="MsoNormal">

<p class="MsoNormal">