Testing Ideological Social Identity’s Effects in Party Primaries Versus General Elections
Mount Vernon Nazarene University
Home page: http://www.mvnu.edu/facstaff/professors/ArtsHuman/cdevine.asp
Sample size: 1250
Field period: 1/17/2010-4/26/2010
This study represents the first empirical measurement of psychological attachment to ideological in-groups, or ideological social identity
(ISI). Building upon the Social Identity Theory literature's theoretical tenets and empirical measurements, I estimate ideological, as well as partisan, social identity levels among a representative participant sample, as well as variation in social identity
levels across ideological groups, electoral contexts, and question order manipulations. My analysis indicates that feelings of psychological attachment to an ideological in-group are common in the mass public, with the average ISI level significantly exceeding
neutral levels. Also, nearly one-third of participants evidence greater relative attachment to ideological, versus partisan, in-groups. Conservatives' have significantly stronger ISI than moderates, while scoring marginally higher than liberals. Also, conservative
Republicans' exhibit significantly greater attachment to their ideological, over partisan, in-group, while liberal Democrats and moderate Independents exhibit greater relative attachment to their partisan in-groups. Exposure to stimuli describing general elections
or party primaries strengthen ISI significantly above baseline (control condition) levels, however ISI does not vary significantly between the two types of electoral contexts. Thus, I conclude that electoral contests, in general, strengthen ISI above baseline
levels. ISI does not vary significantly depending on the order in which participants' answer ideological and partisan social identity measures, however. These results indicate that many individuals incorporate ideological in-groups into their overall social
identities, as extensions of their self-concept. ISI levels do vary across political groups and in response to elections, but they do not vary significantly based on question ordering.
1. What is the average level of ideological social identity in the mass
2. To what extent do ideological social identity levels exceed partisan social identity levels?
3. Do conservatives have higher ideological social identity levels than liberals and moderates?
4. Does ideological social identity become strongest in
party primary contexts, when ideological considerations are most salient? 5. Do reported ideological social identity levels vary depending upon the order in which participants respond to measures of ideological versus partisan social identity?
1. Electoral context. Participants were assigned to one of three conditions. Participants in the first treatment group read a stimulus
page describing a general election between one Democratic and one Republican candidate. Participants in the second treatment group read a stimulus page describing a party primary between one ideologically moderate and one ideologically orthodox (liberal in
a Democratic Party primary or conservative in a Republican Party primary) candidate. Participants in the control condition did not read any stimulus page. 2. Candidate order manipulation. I randomly varied the order in which participants read about either
of the two candidates in the general election and party primary stimuli. 3. Question order manipulation. Participants were randomly assigned to answer ideological social identity measures before partisan social identity measures, or partisan social identity
measures before ideological social identity measures.
Ideological social identity levels. This dependent variable ranges from 1 (weakest ideological social identity) to 7 (strongest
ideological social identity). Scores reflect participants' average level of agreement on a 1 (strongly disagree) to 7 (strongly agree) scale measuring agreement with items tapping psychological attachment to the participant's reported ideological in-group
(liberal, conservative, or moderate). 2. Partisan social identity levels. This dependent variable ranges from 1 (weakest partisan social identity) to 7 (strongest partisan social identity). Scores reflect participants' average level of agreement on a 1 (strongly
disagree) to 7 (strongly agree) scale measuring agreement with items tapping psychological attachment to the participant's reported partisan in-group (Democrat, Republican, or Independent). 3. Relative social identity levels. This outcome variable measures
the difference between each participant's reported ideological and partisan social identity levels. It is calculated by subtracting PSI from ISI scores, so that positive scores represent stronger relative ideological social identity and negative scores represent
stronger relative partisan social identity. Scores range from –6 to +6.
Among 1,089 participants, the average level of ideological social identity – or psychological attachment
to an ideological in-group – is 4.21 (SD = 1.20) on a 1 (weakest social identity) to 7 (strongest social identity) scale. This mean score significantly exceeds the ISI scale's neutral point, t(1079) = 5.68, p = 0.000. Over 41% of participants score above the
scale's neutral point, and 29% average at least slight agreement (5) with ISI measures. On average, the difference between reported ideological and partisan social identity is –0.01. However, 27.4% of participants evidence higher relative attachment to their
ideological, versus partisan, in-group. Comparing across ideological groups, one-way ANOVAs show that conservatives (M = 4.44, SD = 1.25) score significantly higher than moderates on the ISI scale (M = 3.91, SD = 1.07), F (1, 826) = 42.67, p = 0.000, while
conservatives differ from liberals at a marginal level of statistical significance (M = 4.25, SD = 1.21), F(1, 698) = 3.57, p = 0.059). In terms of relative social identity levels, participants identifying as conservatives and Republicans score significantly
higher on ideological social identity (M = 4.62, SD = 1.20) than on partisan social identity (M = 4.40, SD = 1.15), t(350) = 4.19, p = 0.000. In contrast, moderate Independents score significantly higher on PSI (M = 4.10, SD = 1.04) than on ISI (M = 3.60,
SD = 1.01), t(48) = -3.54, p = 0.001. Also, liberal Democrats have a higher PSI (M = 4.46, SD = 1.22) than ISI, but this difference reaches only marginal levels of statistical significance (M = 4.33, SD = 1.22) t(211) = -1.83, p = 0.070. Linear regression
analysis indicates that assignment to the two treatment groups (general election or party primary conditions) significantly increases ideological social identity above baseline (control condition) levels. The regression coefficient is essentially identical
for both treatment group variables (0.260 and 0.256, respectively), indicating that general elections and party primaries strengthen ISI equally. Question order effects are only marginally significant in this analysis, with participants scoring somewhat higher
on the ideological social identity scale when answering ISI before PSI measures.