“There is nothing which I dread so much as a division of the republic into two great parties, each arranged under its leader, and concerting measures in opposition to each other. This, in my humble apprehension, is to be dreaded as the greatest political evil under our Constitution.”
In my last post, I picked on Republicans when looking at the conflict between ingroup favoritism and policy stances. I recommend reading that first (click here) if you want more info, but basically, I found that Republicans were split pretty evenly on support for a program that would require universities to ensure that a particular percentage of employees were Republicans unless I referred to it as “affirmative action” — which led to a significant majority against the idea.
I figured it’s only fair to follow up with a similar survey on Democrats. Whereas Democrats tend to favor affirmative action (according to this Pew Research survey, ~54% of Democrats approve of considering race/ethnicity in college admission decisions), this question is about affirmative action for a political party affiliation — not for a particular racial/ethnic group. The entire justification for accounting for race/ethnicity in college admissions is one of racial justice. America has a bloody history — 245 years of slavery, followed by 100 years of Jim Crow, redlining and the consequences of this history alongside continuing discrimination and terror that persists today. Whereas strong arguments exist for Affirmative action for certain racial/ethic groups, these arguments don’t apply to political party affiliation. Therefore, I would expect that any support for an affirmative action program for Democrats likely stems from ingroup favoritism more than anything else (though I’m open to other perspectives.)
democratic support for Affirmative action in law enforcement
For my last post, I picked academia, because in general, Democrats are more represented in university faculties. For Democrats, I chose law enforcement, since Republicans (again — in general) tend to be more represented among these jobs.
Figure 1 shows a similar story to that of my previous post with Republicans — Democrats are split in their support when the policy benefits them and pretty uniformly against when it supports Republicans.
Figure 1.
An alternate explanation for these results is that respondents know that in reality, Democrats are less well represented in law enforcement and ignored instructions to treat this as a hypothetical (or simply couldn’t overcome belief bias.) To check this, I ran another survey where the target of the policy was Republicans, and the context was faculty positions in academia. This ensures that reality and the hypothetical match (as Republicans tend to be less well represented in university faculties.)
As you can see below, Democrats do not favor an affirmative action type program for Republicans in academia any more than they do in law enforcement.
Figure 2.
Does it matter if it is called affirmative action?
Republican support for a program to ensure a specific percentage of university faculty were Republicans depended on whether the program was referred to as “an affirmative action program”. This makes sense because Affirmative action cues all sorts of negative associations for Republicans. Since it doesn’t have a negative connotation for Democrats, I didn’t expect there to be much difference in terms of support among Democrats. The data largely support this conclusion — Democrats are pretty evenly split in terms of support regardless of whether it is referred to as affirmative action or not. Descriptively, there is a small reduction in support when it’s called Affirmative action — I’m tempted to say that perhaps some Democrats are less comfortable when this term is used (as they realize that the justification for affirmative action isn’t coherent when applied to political parties — but I wouldn’t take this to the bank given how small (and uncertain) the difference is.
Figure 3.
For stats nerds: I ran a Bayesian ordinal regression (with a cumulative link function) using the brms R package (Bürkner, 2017) using default priors. The effect of removing “Affirmative action” wording was b = 0.19 95% CI [-.30, .70], with 77.7% of posterior draws greater than zero (e.g., probability of direction.) (In other words, if I was into betting, I’d probably say that removing the wording increases support, but if I was writing this up for a scientific journal, I’d say we need more data.)
Why this matters.
In my last post, I found that Republicans tended to be split on a program to require that a certain percentage of university faculty were Republicans. Here I’m finding the same pattern for Democrats in a different domain — law enforcement. Whereas Republicans were less supportive when the program was described as “affirmative action” — a term associated with Democratic policy — unsurprisingly, Democratic support didn’t change much based on wording. This results demonstrates that ingroup favoritism isn’t a problem limited to just one party. Anyone belonging to any political party is going to be influenced in some degree by the psychological forces of group membership.
It is important to note (particularly during these times…) that just because any given party is subject to these biases, that all parties are being swayed by these forces to the same degree at all times. Political parties that are built on a diverse coalition of people with a variety of identities and backgrounds are likely be less easily manipulated by party loyalty (as they are coming to the party with many other loyalties.) Parties with strong institutional norms respecting the rule of law (particularly when it doesn’t favor them) are likely to be able to push back on the baser impulses of group identity. Parties that reinforce national identity as superordinate to party identity should also be less likely to ultimately act on these biases.
Up next
I’m planning two more posts on this topic! Next, I’m addressing a potential weakness of my first post on Republicans — that Republicans only care about balancing viewpoints in education, so they shouldn’t show ingroup bias for other areas — by repeating the initial survey, but focusing on law enforcement rather than higher education. The next and final post, I’m going to examine some of the peer-reviewed work in this area and share my current views on the topic of the psychology of political parties.
Like the article? Consider supporting for more like it!
Click here to check out my Patreon page!
Survey details:
All surveys were posted on Cloud research. Participants were limited to those who selected Democrat as party affilation.
Survey 1: “Imagine that a report was released that showed that Democrats are not well represented in certain areas of law enforcement...“
No affirmative action condition: N = 102
Affirmative action condition: N = 98
Total N = 200
Mean age = 37 (sd = 10.96, min = 18, max = 72)
51% male, 48.5% female, 0.5% other
Survey 2: “Imagine that a report was released that showed that Democrats are not well represented in certain areas of law enforcement…“
No affirmative action condition: N = 22
Affirmative action condition: N = 39
Total N = 61*
Mean age = 43 (sd = 16.05, min = 19, max = 99)
34.4% male, 60.7% female, 1.6% Agender, 3.3% no answer
*I only collected 61 because this was not meant to examine the affirmative action vs. no affirmative action wording — it was just meant to show the very clear ingroup bias.