A few weeks ago, Vermont Governor Phil Scott made the announcement that COVID-19 vaccine access in his state was being specifically expanded to all residents who are racial minorities (16 or older):
The disproportionately high toll that the pandemic has taken on communities of color — in terms of infection, hospitalization, and death, and now vaccine access disparities — was cited in support of this decision. The policy was soon met with criticism on social media. Although we still don’t know the contours of public opinion on this policy, one reason that might motivate opposition is self-interest: the majority of the public is white and thus does not qualify for this expanded vaccine access, and so supporting this policy presents a cost for white Americans (especially those who are not already vaccinated).
At the same time, some whites might want to help redress racial inequities in society more broadly as well as in specific cases like the pandemic. Indeed, arguably the biggest recent public opinion trend has been the increasing racial liberalism among whites (especially Democrats). This vaccine expansion context offers an interesting test of this growing racial liberalism — does it compel whites to support a policy like race-targeted vaccine expansion that creates a cost in their lives? If so, then maybe the growing racial liberalism among whites (and endorsing ideas like racial inequalities are a problem in society that ought to be solved) is sincere and meaningful.
I looked into this question in an early April Lucid survey of about 700 white Americans nationwide (out of about 900 total respondents, after filtering out inattentive survey-takers). Early on in this survey, respondents where asked to respond to two of the four standard racial resentment items:
- Irish, Italians, Jewish and many other minorities overcame prejudice and worked their way up. Blacks should do the same without any special favors.
- Generations of slavery and discrimination have created conditions that make it difficult for blacks to work their way out of the lower class.
Responses were on a 1-5 strength of agreement scale, and recoded to a 0-1 scale where higher values indicate higher levels of racial resentment. I then averaged the two items together and created a four-category racial resentment level variable based on the average: Low (0-0.25), Low Middle (0.25-0.5), High Middle (0.5-0.75), and High (0.75-1). This helps with data presentation and allows me to not assume a linear relationship between racial resentment and an outcome of interest (results are similar if I treat this as a continuous variable though).
Later on in the survey, respondents gave their opinions on the type of policy that the Vermont governor introduced (adapted for a national sample):
- Some states have recently expanded access to the COVID-19 vaccine to all individuals who identify as “Black, Indigenous, or a person of color” who are 16 years or older. Do you support or oppose this expansion of vaccine access?
Responses here were also on a 1-5 strength of agreement scale. For purposes of easy presentation, I recode this variable to equal 1 if respondents strongly or somewhat support the policy, and 0 otherwise. The below plot shows policy support by the four-category racial resentment variable:
The plot shows that as racial resentment decreases — going from right to left — support for race-targeted vaccine expansion goes up. Whites in the highest racial resentment category support the policy at 43% while those in the lowest group sit at 91%.
Circling back to the earlier motivating question for this analysis, evidence here suggests that those on the lowest end of the scale — racial liberals, a group that’s been growing in size as of late — are actually acting on those beliefs in supporting a policy that may come as a cost to themselves but a benefit towards addressing racial inequalities in vaccine access.
Caveats and Other Things to Consider
- All data analysis comes with limitations, and there’s an especially big caveat here: white respondents may not have correctly perceived that the policy — given the description I gave — would come at a cost to them (i.e. understanding that vaccine expansion to people of color would occur before expansion to all whites). In future questions of this sort, it will be important to make this point clear in the question text, for example. I don’t think this element would change the slope of the relationship between racial resentment and policy support, but it almost certainly will decrease support levels across all respondents. Along with the fact that this is not a nationally representative sample — and in general, Lucid makes more for more liberal samples than average — true levels of support for a policy like this will be lower than shown here (overall and among all racial resentment level groups). But again, that doesn’t necessarily change the takeaway that racial resentment exerts a powerful influence on opinions regarding this policy (and that racial liberals are likely acting on their views at higher than average levels).
- Three types of patterns can produce the relationship above: whites who are low on the racial resentment scale supporting the policy, whites who are high on racial resentment opposing the policy, or a mix of both of these. All three explanations for why a relationship appears are observationally equivalent — this empirical setup does not give us any insight into what’s driving what the results show. (This is an argument that co-authors and I make in a recent paper; randomizing minority vs. non-minority targets as the beneficiaries of vaccine expansion in a survey experiment, for example, would help disentangle these different patterns.)
- Costs to supporting this policy would be highest among whites who have not already received the COVID vaccine. Unfortunately, a measure of this was not included in the survey. One strong correlate of vaccine receipt is age. The racial resentment/policy support relationship and overall supports levels are very similar among whites below 50 years or old vs. at 50 or older however. For example, for moving from Low to High racial resentment groups, support rate falls from 93% to 41% among whites below 50 and from 90% to 44% among whites 50 or older.