Background

We are interested in the effect of class- vs. race-based societal zero-sum thinking on perceived impact of economic policy and support for that policy. To that end, we measure societal zero-sum beliefs of both race and class, introduce progressive policy (minimum wage OR student debt relief OR marijuana legislation), ask about this policy’s impact on racial/economic equality, working class white people, upper class white people, and working class black people.

Unfortunately, I F’ed the black upper class impact question. It was mislabeled (participants brought this to my attention in the open feedback question at the end), so that measure is unusable.

There are three main things that are different in this study, compared to the last one:
1. Class-based status zero-sum was reverse-framed from last time. That is, instead of framing it as “working class benefitting at the expense of upper class” (last study), we framed it as “upper class benefitting at the expense of working class.”
2. We changed the impact DV’s: last time we asked how the policy impacts all white people, all black people, all working class people, and all upper class people. This time, we broke it down: how does this policy impact the average white working class person, the average white upper class person, the average black working class person, and the average black working class person (the last one wasn’t measured due to an error on my end). The idea of this is to allow us to see differential impact on race vs. class as predicted by zero-sum beliefs.
3. We added resource zero-sum beliefs as an additional measure. We have status/power zero-sum beliefs, but resources might behave a little differently, so we added it as an epxloratory IV.

Demographics

Race

race N Perc
American Indian or Alaska Native 1 0.50
Asian 14 6.97
Black or African American 27 13.43
Hispanic, Latino, or Spanish origin 2 1.00
Middle Eastern or North African 1 0.50
Native Hawaiian or Other Pacific Islander 1 0.50
White 145 72.14
multiracial 9 4.48
NA 1 0.50

Gender

gender N Perc
man 124 61.69
woman 73 36.32
NA 4 1.99

Age

age_mean age_sd
38.21 11.55

Education

edu N Perc
noHS 1 0.50
GED 56 27.86
2yearColl 20 9.95
4yearColl 81 40.30
MA 34 16.92
PHD 6 2.99
NA 3 1.49

Class

class N Perc
Lower Class 14 6.97
Working Class 43 21.39
Lower Middle Class 87 43.28
Upper Middle Class 56 27.86
Upper Class 1 0.50

Income

County-level data

We also asked them which county they live in. And then, with census data, we got their county’s GINI coefficient, median income, and population density.

GINI

Median income

Population density

Politics

Ideology

Participants were asked about the extent to which they subscribe to the following ideologies on a scale of 1-7 (select NA if unfamiliar): Conservatism, Liberalism, Democratic Socialism, Libertarianism, Progressivism, Right-Wing Nationalism (removed from plot because it’s too heavily skewed).

Party ID

party_id N Perc
Democrat 109 54.23
Independent 48 23.88
Republican 44 21.89

Vote in 2020

vote_2020 N Perc
Joe Biden 103 51.24
Donald Trump 46 22.89
I did not vote 43 21.39
Third-party candidate 9 4.48

Vote in 2024

vote_2024 N Perc
Joe Biden 95 47.26
Donald Trump 45 22.39
I will not vote 35 17.41
Other 18 8.96
Cornel West 5 2.49
Robert F. Kennedy Jr.  3 1.49

Measures

Class-based Status Zero-Sum Beliefs

Adapted from Davidai & Ongis, 2019: https://www.science.org/doi/pdf/10.1126/sciadv.aay3761

  1. The more people in the upper class obtain positions of power, the harder it is for people in the working class to attain positions of power in society
  2. The easier it is for upper class students to gain admission to college, the more difficult it becomes for working class students to get admitted to college
  3. The more resources the government spends on predominantly upper class regions in the U.S., the less it spends on predominantly working class regions
  4. The more influence the upper class has in politics, the less influence the working class has in politics
  5. When people in the upper class move up in society, they do so at the expense of people in the working class
  6. The easier it is for people in the upper class to get high-paying jobs, the more difficult it becomes for people in the working class to obtain similar high-paying jobs

    alpha = 0.9

Race-based Status Zero-Sum Beliefs

Adapted from Davidai & Ongis, 2019: https://www.science.org/doi/pdf/10.1126/sciadv.aay3761

  1. The more racial minorities obtain positions of power, the harder it is for white people to attain positions of power in society
  2. The easier it is for racial minorities to gain admission to college, the more difficult it becomes for white students to get admitted to college
  3. The more resources the government spends on predominantly racial minority regions in the U.S., the less it spends on predominantly white regions
  4. The more influence racial minorities have in politics, the less influence white people have in politics
  5. When racial minorities move up in society, they do so at the expense of white people
  6. The easier it is for racial minorities to get high-paying jobs, the more difficult it becomes for white people to obtain similar high-paying jobs

    alpha = 0.93

Class-based Resource Zero-Sum Beliefs

Adapted from Chinoy et al., 2022: https://nathannunn.sites.olt.ubc.ca/files/2022/12/Zero_Sum_US_Political_Divides.pdf

  1. If the upper class becomes richer, this comes at the expense of the working class
  2. If the upper class makes more money, then the working class makes less money
  3. If the upper class does better economically, this does NOT come at the expense of the working class [R]

    alpha = 0.9

Race-based Resource Zero-Sum Beliefs

Adapted from Chinoy et al., 2022: https://nathannunn.sites.olt.ubc.ca/files/2022/12/Zero_Sum_US_Political_Divides.pdf

  1. If racial minorities become richer, this comes at the expense of white people
  2. If racial minorities make more money, then white people make less money
  3. If racial minorities do better economically, this does NOT come at the expense of white people [R]

    alpha = 0.83

Policy

Participants were shown one of the following policies.

Minimum wage: Congress has not increased the federal minimum wage, currently set at 7.25 Dollars, since 2009. Some Congresspeople are proposing a policy that would gradually raise the federal minimum wage to 15 Dollars an hour by 2025. After 2025, the minimum wage would be adjusted each year to keep pace with growth in the median wage, a measure of wages for typical workers.

Student debt forgiveness: Some Congresspeople are proposing a policy that would help to address the student loan debt crisis by forgiving up to 50,000 Dollars in loans per borrower. Approximately 42 million Americans, or about 1 in 6 American adults, owe a cumulative 1.6 trillion Dollars in student loans. Student loans are now the second-largest slice of household debt after mortgages, bigger than credit card debt.

Marijuana legislation: Some Congresspeople are proposing a bill which would lead to comprehensive marijuana reform legislation. It would help rebuild communities that have been the most impacted by marijuana criminalization under federal law. It would also remove marijuana from the federal list of banned substances, allowing states to set their own marijuana policies. It would also create pathways to erase previous convictions and prohibit discrimination based on the use or possession of marijuana.

Policy support

To what extent do you oppose or support this policy? (1 = Strongly Oppose to 7 = Strongly Support)

Impact on racial/economic equality

How do you think this policy will impact… (-3 = Greatly reduce equality to 3 = Greatly increase equality)
Racial equality is the equal distribution of resources, power, and economic opportunity across racial groups in society.
Economic equality refers to the extent to which income and wealth are evenly distributed in the US.

1. Racial equality
2. Economic equality

Impact on groups

In your opinion, what is the impact of this policy on… (-3 = Extremely Harmful to 3 = Extremely Helpful)

wwc: The average white working class person?
wuc: The average white upper class person?
bwc: The average black working class person?

Analysis

Correlation Matrix

In general, looks like resource is doing more for class zs and status is doing more for race zs. It’s interesting, but it’s also a bit of a methodological problem.

DV: Support for Policy

IV: status race zs

(#tab:unnamed-chunk-27)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 6.75 [6.14, 7.35] 21.99 199 < .001
Zs status race -0.42 [-0.60, -0.24] -4.62 199 < .001

IV: status class zs

(#tab:unnamed-chunk-28)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 4.29 [3.27, 5.31] 8.28 199 < .001
Zs status class 0.23 [0.04, 0.43] 2.33 199 .021

Opposite effects…

IV: resource race zs

(#tab:unnamed-chunk-29)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 6.39 [5.78, 6.99] 20.86 199 < .001
Zs resource race -0.35 [-0.55, -0.14] -3.35 199 < .001

IV: resource class zs

(#tab:unnamed-chunk-30)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 3.94 [3.15, 4.74] 9.74 199 < .001
Zs resource class 0.33 [0.16, 0.49] 3.94 199 < .001

ok, this is very interesting, actually. Let’s keep going.

IV: resource class zs and resource race zs

(#tab:unnamed-chunk-31)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 4.85 [3.95, 5.75] 10.62 198 < .001
Zs resource class 0.35 [0.19, 0.51] 4.38 198 < .001
Zs resource race -0.38 [-0.58, -0.19] -3.85 198 < .001

IV: Interaction term of resource class vs. resource race

I’m gonna try something here. To try and compare the effect of race vs. class zs, I’ll create a within-subject variable called type (race vs. class) and a continuous variable called zs_resource. Then, I’ll interact the two to predict support. Let’s see if this makes sense.

(#tab:unnamed-chunk-33)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 3.94 [3.14, 4.74] 9.69 398 < .001
Zs resource 0.33 [0.16, 0.49] 3.92 398 < .001
Typerace 2.44 [1.44, 3.44] 4.80 398 < .001
Zs resource \(\times\) Typerace -0.67 [-0.93, -0.41] -5.08 398 < .001

let’s visualize.

Is this kind of wild or is it just an artifact of us messing with the framing/directionality of the zero-sum?

Let’s see how the status version of this looks.

IV: Interaction term of status class vs. status race

(#tab:unnamed-chunk-36)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 4.29 [3.29, 5.29] 8.44 398 < .001
Zs status 0.23 [0.04, 0.43] 2.37 398 .018
Typerace 2.46 [1.28, 3.63] 4.11 398 < .001
Zs status \(\times\) Typerace -0.65 [-0.92, -0.39] -4.83 398 < .001

let’s visualize.

Yeah. This tracks. Alright, let’s keep going.

Controlling for a bunch of things

IV: resource class

(#tab:unnamed-chunk-38)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 5.28 [4.30, 6.25] 10.66 187 < .001
Zs resource class 0.23 [0.07, 0.40] 2.80 187 .006
Ideo con -0.33 [-0.46, -0.20] -5.06 187 < .001
(#tab:unnamed-chunk-39)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 6.63 [2.77, 10.48] 3.39 169 < .001
Zs resource class 0.24 [0.06, 0.42] 2.68 169 .008
Ideo con -0.31 [-0.45, -0.18] -4.50 169 < .001
As numericincome 0.01 [-0.12, 0.13] 0.10 169 .920
As numericedu -0.06 [-0.32, 0.20] -0.46 169 .648
As numericclass 0.03 [-0.35, 0.41] 0.17 169 .868
RaceAsian -2.88 [-6.60, 0.84] -1.53 169 .128
RaceBlack or African American -1.10 [-4.68, 2.48] -0.60 169 .546
RaceHispanic, Latino, or Spanish origin -2.37 [-6.67, 1.93] -1.09 169 .278
RaceMiddle Eastern or North African -1.09 [-6.02, 3.85] -0.43 169 .665
Racemultiracial -1.56 [-5.29, 2.18] -0.82 169 .411
RaceNative Hawaiian or Other Pacific Islander -1.51 [-6.46, 3.44] -0.60 169 .548
RaceWhite -1.58 [-5.10, 1.94] -0.88 169 .378
Genderwoman 0.56 [0.00, 1.11] 1.99 169 .048
(#tab:unnamed-chunk-40)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 7.45 [3.49, 11.40] 3.72 155 < .001
Zs resource class 0.24 [0.06, 0.43] 2.56 155 .011
Ideo con -0.34 [-0.48, -0.20] -4.65 155 < .001
As numericincome 0.01 [-0.12, 0.14] 0.15 155 .879
As numericedu -0.16 [-0.43, 0.12] -1.14 155 .258
As numericclass 0.03 [-0.36, 0.42] 0.16 155 .876
RaceAsian -3.30 [-7.07, 0.47] -1.73 155 .086
RaceBlack or African American -1.52 [-5.14, 2.10] -0.83 155 .407
RaceHispanic, Latino, or Spanish origin -2.71 [-7.04, 1.61] -1.24 155 .217
RaceMiddle Eastern or North African -1.33 [-6.30, 3.64] -0.53 155 .597
Racemultiracial -1.89 [-5.73, 1.94] -0.97 155 .331
RaceNative Hawaiian or Other Pacific Islander -1.88 [-6.86, 3.09] -0.75 155 .456
RaceWhite -2.07 [-5.64, 1.51] -1.14 155 .256
Genderwoman 0.54 [-0.04, 1.12] 1.83 155 .068
Scalecounty gini 0.13 [-0.22, 0.49] 0.74 155 .459
Scalecounty density 0.00 [-0.37, 0.37] 0.00 155 .997
Scalecounty medianincome 0.11 [-0.18, 0.40] 0.75 155 .453

it holds…

IV: resource race

(#tab:unnamed-chunk-41)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 6.83 [6.21, 7.46] 21.60 187 < .001
Zs resource race -0.16 [-0.37, 0.05] -1.48 187 .141
Ideo con -0.35 [-0.48, -0.21] -5.06 187 < .001

oh wow, conservatism sucks up all the variance from resource zero-sum race.

(#tab:unnamed-chunk-42)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 7.99 [4.14, 11.84] 4.10 169 < .001
Zs resource race -0.19 [-0.42, 0.03] -1.68 169 .095
Ideo con -0.33 [-0.47, -0.19] -4.59 169 < .001
As numericincome -0.03 [-0.16, 0.10] -0.51 169 .609
As numericedu 0.02 [-0.24, 0.28] 0.13 169 .898
As numericclass 0.02 [-0.36, 0.40] 0.10 169 .919
RaceAsian -2.58 [-6.33, 1.18] -1.35 169 .178
RaceBlack or African American -0.95 [-4.57, 2.67] -0.52 169 .605
RaceHispanic, Latino, or Spanish origin -2.00 [-6.35, 2.34] -0.91 169 .364
RaceMiddle Eastern or North African -0.98 [-5.98, 4.02] -0.39 169 .700
Racemultiracial -1.62 [-5.40, 2.17] -0.84 169 .400
RaceNative Hawaiian or Other Pacific Islander -1.09 [-6.09, 3.91] -0.43 169 .668
RaceWhite -1.32 [-4.88, 2.24] -0.73 169 .466
Genderwoman 0.68 [0.12, 1.24] 2.40 169 .018
(#tab:unnamed-chunk-43)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 8.85 [4.93, 12.78] 4.46 155 < .001
Zs resource race -0.23 [-0.46, 0.01] -1.90 155 .059
Ideo con -0.34 [-0.49, -0.20] -4.62 155 < .001
As numericincome -0.04 [-0.18, 0.10] -0.59 155 .558
As numericedu -0.08 [-0.36, 0.20] -0.57 155 .568
As numericclass 0.06 [-0.34, 0.45] 0.28 155 .783
RaceAsian -3.09 [-6.89, 0.70] -1.61 155 .110
RaceBlack or African American -1.39 [-5.04, 2.26] -0.75 155 .454
RaceHispanic, Latino, or Spanish origin -2.39 [-6.75, 1.97] -1.08 155 .281
RaceMiddle Eastern or North African -1.38 [-6.40, 3.64] -0.54 155 .588
Racemultiracial -1.84 [-5.72, 2.03] -0.94 155 .349
RaceNative Hawaiian or Other Pacific Islander -1.59 [-6.59, 3.42] -0.63 155 .532
RaceWhite -1.86 [-5.47, 1.75] -1.02 155 .309
Genderwoman 0.71 [0.12, 1.29] 2.39 155 .018
Scalecounty gini 0.15 [-0.21, 0.51] 0.83 155 .406
Scalecounty density -0.05 [-0.42, 0.32] -0.26 155 .796
Scalecounty medianincome 0.19 [-0.11, 0.49] 1.26 155 .210

IV: status class

(#tab:unnamed-chunk-44)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 5.49 [4.40, 6.58] 9.93 187 < .001
Zs status class 0.20 [0.01, 0.39] 2.03 187 .043
Ideo con -0.37 [-0.49, -0.24] -5.72 187 < .001

barely.

(#tab:unnamed-chunk-45)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 6.76 [2.86, 10.67] 3.42 169 < .001
Zs status class 0.21 [0.01, 0.41] 2.06 169 .041
Ideo con -0.35 [-0.48, -0.22] -5.16 169 < .001
As numericincome 0.01 [-0.12, 0.14] 0.16 169 .871
As numericedu -0.04 [-0.29, 0.22] -0.29 169 .776
As numericclass -0.01 [-0.39, 0.37] -0.05 169 .962
RaceAsian -2.91 [-6.67, 0.85] -1.53 169 .128
RaceBlack or African American -1.10 [-4.71, 2.52] -0.60 169 .550
RaceHispanic, Latino, or Spanish origin -2.25 [-6.59, 2.08] -1.02 169 .307
RaceMiddle Eastern or North African -1.11 [-6.09, 3.87] -0.44 169 .661
Racemultiracial -1.76 [-5.53, 2.02] -0.92 169 .359
RaceNative Hawaiian or Other Pacific Islander -1.23 [-6.22, 3.75] -0.49 169 .626
RaceWhite -1.52 [-5.08, 2.03] -0.85 169 .399
Genderwoman 0.64 [0.08, 1.19] 2.26 169 .025
(#tab:unnamed-chunk-46)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 7.62 [3.61, 11.62] 3.76 155 < .001
Zs status class 0.21 [0.00, 0.41] 1.96 155 .052
Ideo con -0.37 [-0.51, -0.23] -5.19 155 < .001
As numericincome 0.01 [-0.12, 0.15] 0.19 155 .848
As numericedu -0.14 [-0.42, 0.13] -1.01 155 .313
As numericclass 0.00 [-0.39, 0.39] 0.00 155 .997
RaceAsian -3.38 [-7.19, 0.43] -1.75 155 .082
RaceBlack or African American -1.55 [-5.20, 2.10] -0.84 155 .403
RaceHispanic, Latino, or Spanish origin -2.60 [-6.96, 1.76] -1.18 155 .241
RaceMiddle Eastern or North African -1.40 [-6.42, 3.62] -0.55 155 .582
Racemultiracial -2.06 [-5.94, 1.82] -1.05 155 .296
RaceNative Hawaiian or Other Pacific Islander -1.62 [-6.63, 3.38] -0.64 155 .523
RaceWhite -2.04 [-5.65, 1.57] -1.12 155 .266
Genderwoman 0.64 [0.06, 1.22] 2.19 155 .030
Scalecounty gini 0.17 [-0.19, 0.53] 0.93 155 .355
Scalecounty density -0.01 [-0.38, 0.36] -0.07 155 .947
Scalecounty medianincome 0.14 [-0.15, 0.43] 0.94 155 .350

IV: status race

(#tab:unnamed-chunk-47)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 7.01 [6.39, 7.63] 22.26 187 < .001
Zs status race -0.22 [-0.42, -0.03] -2.29 187 .023
Ideo con -0.31 [-0.45, -0.18] -4.48 187 < .001
(#tab:unnamed-chunk-48)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 7.83 [4.02, 11.65] 4.05 169 < .001
Zs status race -0.22 [-0.43, -0.01] -2.07 169 .040
Ideo con -0.31 [-0.45, -0.16] -4.18 169 < .001
As numericincome -0.04 [-0.17, 0.09] -0.56 169 .579
As numericedu 0.01 [-0.24, 0.27] 0.10 169 .919
As numericclass 0.07 [-0.32, 0.46] 0.36 169 .722
RaceAsian -2.44 [-6.18, 1.30] -1.29 169 .199
RaceBlack or African American -0.79 [-4.40, 2.81] -0.43 169 .665
RaceHispanic, Latino, or Spanish origin -1.73 [-6.06, 2.60] -0.79 169 .431
RaceMiddle Eastern or North African -0.63 [-5.60, 4.35] -0.25 169 .804
Racemultiracial -1.36 [-5.13, 2.41] -0.71 169 .476
RaceNative Hawaiian or Other Pacific Islander -0.99 [-5.95, 3.98] -0.39 169 .696
RaceWhite -1.17 [-4.72, 2.37] -0.65 169 .515
Genderwoman 0.61 [0.05, 1.16] 2.16 169 .032
(#tab:unnamed-chunk-49)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 8.63 [4.75, 12.51] 4.40 155 < .001
Zs status race -0.27 [-0.48, -0.05] -2.46 155 .015
Ideo con -0.32 [-0.47, -0.17] -4.29 155 < .001
As numericincome -0.04 [-0.17, 0.10] -0.58 155 .565
As numericedu -0.09 [-0.36, 0.19] -0.63 155 .530
As numericclass 0.12 [-0.28, 0.52] 0.60 155 .548
RaceAsian -2.88 [-6.65, 0.89] -1.51 155 .134
RaceBlack or African American -1.18 [-4.80, 2.44] -0.64 155 .521
RaceHispanic, Latino, or Spanish origin -2.02 [-6.35, 2.31] -0.92 155 .358
RaceMiddle Eastern or North African -0.89 [-5.87, 4.08] -0.35 155 .724
Racemultiracial -1.45 [-5.30, 2.41] -0.74 155 .460
RaceNative Hawaiian or Other Pacific Islander -1.44 [-6.39, 3.52] -0.57 155 .568
RaceWhite -1.65 [-5.23, 1.94] -0.91 155 .365
Genderwoman 0.62 [0.04, 1.20] 2.13 155 .035
Scalecounty gini 0.16 [-0.20, 0.52] 0.89 155 .375
Scalecounty density -0.04 [-0.40, 0.33] -0.19 155 .846
Scalecounty medianincome 0.14 [-0.15, 0.43] 0.96 155 .338

yeah, status does a lot more for race zs, and resource does a lot more for class zs. I feel like this is a paper in itself. For us it’s a bit of a challenge in terms of measurement, but, I mean, this challenge tells us a lot about america, no?

DV: Impact of policy on racial equality

IV: status race zs

(#tab:unnamed-chunk-50)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 1.24 [0.84, 1.63] 6.12 199 < .001
Zs status race -0.13 [-0.25, -0.02] -2.26 199 .025

IV: status class zs

(#tab:unnamed-chunk-51)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 0.14 [-0.51, 0.79] 0.42 199 .675
Zs status class 0.14 [0.01, 0.26] 2.15 199 .033

IV: resource race zs

(#tab:unnamed-chunk-52)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 0.95 [0.56, 1.34] 4.77 199 < .001
Zs resource race -0.05 [-0.18, 0.09] -0.71 199 .476

IV: resource class zs

(#tab:unnamed-chunk-53)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 0.03 [-0.49, 0.54] 0.10 199 .920
Zs resource class 0.17 [0.07, 0.28] 3.23 199 .001

hmm ok.

DV: Impact of policy on economic equality

IV: status race zs

(#tab:unnamed-chunk-54)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 1.71 [1.27, 2.16] 7.57 199 < .001
Zs status race -0.19 [-0.33, -0.06] -2.92 199 .004

IV: status class zs

(#tab:unnamed-chunk-55)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 0.75 [0.01, 1.49] 2.00 199 .047
Zs status class 0.07 [-0.07, 0.22] 1.01 199 .316

IV: resource race zs

(#tab:unnamed-chunk-56)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 1.36 [0.92, 1.80] 6.05 199 < .001
Zs resource race -0.09 [-0.24, 0.06] -1.20 199 .230

IV: resource class zs

(#tab:unnamed-chunk-57)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 0.43 [-0.15, 1.02] 1.46 199 .145
Zs resource class 0.15 [0.03, 0.27] 2.42 199 .017

tracks..

Interaction of zs and equality type

Let’s try something.

and for status zs…

All of this looks pretty similar to the support DV. I guess that’s good: people want equality. I wonder if equality mediates these relationships…

Mediation model 1

status race -> impact on racial equality -> support for policy

Mediation model 2

status class -> impact on economic equality -> support for policy

a = 0.07 (p = 0.316)
b = 0.62 (p = 0)
direct = 0.23 (p = 0.021)
indirect = 0.19 (p = 0.038)

## Mediation model 3

resource race -> impact on racial equality -> support for policy

Mediation model 4

resource class -> impact on economic equality -> support for policy

a = 0.15 (p = 0.017)
b = 0.59 (p = 0)
direct = 0.33 (p = 0)
indirect = 0.24 (p = 0.002)


hmm yeah. looks like resource is doing the most for class and status is doing the most for race… again.

DV: Impact on different groups

###: Impact on white upper class

(#tab:unnamed-chunk-64)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 0.64 [0.10, 1.19] 2.34 199 .020
Zs resource class -0.06 [-0.17, 0.06] -0.98 199 .327
(#tab:unnamed-chunk-65)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 0.26 [-0.15, 0.66] 1.24 199 .215
Zs resource race 0.05 [-0.09, 0.19] 0.70 199 .483
(#tab:unnamed-chunk-66)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 0.50 [-0.18, 1.18] 1.44 199 .153
Zs status class -0.02 [-0.15, 0.11] -0.32 199 .748
(#tab:unnamed-chunk-67)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 0.24 [-0.18, 0.66] 1.13 199 .259
Zs status race 0.05 [-0.08, 0.17] 0.77 199 .443

ok, none of these seem to affect perceived impact of these policies on white upper class.

###: Impact on white working class

(#tab:unnamed-chunk-68)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 0.97 [0.35, 1.58] 3.12 199 .002
Zs resource class 0.07 [-0.05, 0.20] 1.14 199 .257
(#tab:unnamed-chunk-69)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 1.79 [1.34, 2.24] 7.81 199 < .001
Zs resource race -0.18 [-0.34, -0.03] -2.37 199 .019

hmm, the more you think status race is zero-sum -> the more you think these policies will hurt white working class. ok.

(#tab:unnamed-chunk-70)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 0.73 [-0.03, 1.49] 1.90 199 .059
Zs status class 0.11 [-0.03, 0.26] 1.52 199 .130
(#tab:unnamed-chunk-71)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 1.87 [1.41, 2.34] 8.00 199 < .001
Zs status race -0.19 [-0.32, -0.05] -2.70 199 .007

same is true for resource zero-sum. And no effect for class. Very interesting.

###: Impact on black working class

(#tab:unnamed-chunk-72)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 1.01 [0.40, 1.62] 3.26 199 .001
Zs resource class 0.10 [-0.03, 0.22] 1.52 199 .130
(#tab:unnamed-chunk-73)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 2.00 [1.55, 2.45] 8.71 199 < .001
Zs resource race -0.20 [-0.36, -0.05] -2.61 199 .010
(#tab:unnamed-chunk-74)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 1.01 [0.25, 1.78] 2.61 199 .010
Zs status class 0.09 [-0.06, 0.24] 1.18 199 .238
(#tab:unnamed-chunk-75)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 2.22 [1.76, 2.67] 9.54 199 < .001
Zs status race -0.25 [-0.38, -0.11] -3.59 199 < .001

hmm, but high race zsb’s believe that these policies are also hurting black working class - not just white working class. that’s strange. Maybe this is all explained by ideology?

with ideology as a control variable

(#tab:unnamed-chunk-76)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 1.96 [1.47, 2.46] 7.84 187 < .001
Zs resource race -0.12 [-0.29, 0.05] -1.37 187 .171
Ideo con -0.12 [-0.23, -0.01] -2.25 187 .026
(#tab:unnamed-chunk-77)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 1.99 [1.50, 2.49] 7.94 187 < .001
Zs status race -0.12 [-0.27, 0.03] -1.54 187 .125
Ideo con -0.11 [-0.22, 0.00] -1.99 187 .048
(#tab:unnamed-chunk-78)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 2.26 [1.77, 2.74] 9.19 187 < .001
Zs resource race -0.10 [-0.27, 0.06] -1.22 187 .224
Ideo con -0.19 [-0.29, -0.08] -3.50 187 < .001
(#tab:unnamed-chunk-79)
Predictor \(b\) 95% CI \(t\) \(\mathit{df}\) \(p\)
Intercept 2.38 [1.90, 2.86] 9.72 187 < .001
Zs status race -0.15 [-0.30, 0.00] -1.94 187 .054
Ideo con -0.16 [-0.27, -0.06] -3.00 187 .003

yeah…