Age-Based Vaccine Eligibility Ensures Racial Bias
Public health officials, reporters, and local leaders have increasingly settled on the narrative of “hesitancy” as the reason BIPOC communities have not been vaccinated at rates equal to their share of the population. They point to well-documented racial biases in the healthcare system and polling from during the Trump Administration to explain why BIPOC communities are not getting vaccinated at the same levels as Whites. Biases in our health system are very real and some people do not trust the vaccine, but that is not the reason for the racial skew we have been seeing. The problem is the CDC’s policy of age-based vaccine eligibility.
Compared to their overall population, BIPOC communities make up a disproportionately small part of the 65+ population that is eligible for the vaccine.
As the chart to the left shows, CDC’s decision to make age the primary eligibility qualification ensures a dramatically skewed outcome in vaccination rates by race/ethnicity, even if there is perfect compliance and equal participation.
Furthermore, when reported vaccination rates are measured against eligible populations, BIPOC are actually getting vaccinated at rates equal to and above expected levels.
We need to understand what is really going on — both to fix the current vaccine policy and avoid perpetuating a dangerous narrative that disempowers the BIPOC community.
Failure to Understand Age is Creating a Harmful Narrative
Reported vaccination rates based on eligible population — as opposed to total population — show a very different outcome than the one making headlines. Early data consistently indicates that BIPOC people who are eligible are getting vaccinated at rates equal or exceeding their share of the population.
The graph above shows how BIPOC vaccination rates exceed their proportion of the eligible 65+ population. This indicates that BIPOC communities are vaccinating as they are able and reinforces the need to eliminate age-based eligibility criteria in order for BIPOC vaccination rates to match their proportion of the overall population. (Source: US Census and CDC).
Does this mean the system is working? No. The myriad of disconnected and cumbersome local systems are well documented, and still leave BIPOC communities underrepresented compared to total populations and those infected/killed by COVID. The data being reported is also badly incomplete (more on that below). But the data we do have points to a very different narrative of what is happening than the one of hesitancy that has been portrayed in the mainstream media.
BIPOC communities have largely been overcoming the documented barriers to access that have been notoriously challenging for low-income and under-served communities.
Rather than BIPOC communities not being vaccinated because they aren’t willing to accept the cure or can’t navigate technical systems, BIPOC communities have largely been overcoming the documented barriers to access that have been notoriously challenging for low-income and under-served communities. The media’s narrative that BIPOC communities are widely hesitant risks undermining this achievement.
One of the surest ways to drive down participation is to tell a community that most people like them don’t trust the vaccine and that they won’t be able to get it if they try.
So while the U.S. does need to fix broken systems and the eligibility biases that are skewing racial access to the vaccine, we also need to tell a more complete story about the desire to be vaccinated by BIPOC communities and the lengths to which they have gone to overcome the barriers that do exist.
In order to tell a better and truer narrative about what is happening, we need to start with a more nuanced understanding of the data and pay attention to what is missing in it.
The Problem with Mainstream Data Interpretations
Along with missing the importance of distinguishing between eligible and total populations, the media and public health experts have been comparing vaccination demographic data to U.S. Census demographic data — which is an apples to oranges comparison given the very different ways they count race and ethnicity.
The U.S. Census groups all people in distinct categories, whereas vaccine data gives people the option of identifying as “Other” or “Refuse to Answer.” “Multi-race” has also skewed the data set, with vaccine data heavily over-counting this group compared to the U.S. Census. For many states, these categories have garnered as much as 20–30% of the population, adding significant holes to the data when it comes to understanding unique racial/ethnic representation. This is especially true for Blacks and Hispanics, who are more likely to identify as “Other” or “Multi-Race” or “Refuse to Answer.”
Because it is impossible to assign “Other” and “Refuse to Answer” to one of the predominant Census categories, and because public health officials and the media recognize the problem with “Multi-Race” over-count, most often these answers are simply left out of reported totals. These same reports then directly compare incomplete vaccination data to more complete Census data without acknowledging what is concealed by these reporting discrepancies. Often, they do not even recalculate the denominator! For example, in a vaccine cohort reporting as 35% Black, 35% White, and 30% Refused, press coverage has focused on how “only 35% of those vaccinated are Black.” But in reality half of those with known race are Black, and then we do not have data on a third of the population . This is why in many cases, reported White vaccination rates also lag below population levels.
California as Case Study
California’s vaccine data helps bring the problem, and the solution, into focus.
California has much better overall racial representation in its vaccinated population than other states, due in large part to its expanded eligibility criteria. But the way California is reporting race in its vaccine data has largely obscured this achievement.
California’s “Other”, “Multi-Race”, and “Refuse to Answer/Unknown” groups make up 33.9% of the vaccinated population. As seen in the graph below, these groups are much larger than reported in the U.S. Census.
Data as of 2/25/21
This overrepresentation of non-census groups and the inclusion of “Refuse to Answer” has caused underrepresentation in every single race (with those most likely to choose “Multi” and “Other” most undercounted). Yet when we look at White and BIPOC as a whole, we see that the BIPOC community is well-represented in the reported data.
Data as of 2/25/21 with “Unknown/Refused” redistributed proportionately by race.
We need a more honest narrative about BIPOC attitudes toward the COVID vaccines and the reasons these communities have not yet been vaccinated at representative levels. And we need solutions that are realistic and can be quickly implemented to mitigate the bias in current age-based criteria.
A solution that can be immediately implemented nation-wide is for CDC to recommend that states expand eligibility to everyone who lives (or works) in ZIP codes with high COVID-19 infection and death rates. States already collect and report on infection and death by ZIP or county, and they collect information on where vaccinated people live. States can readily change eligibility criteria to allow people who live in the areas at greatest risk of infection and death to get in line for our rapidly expanding vaccine supply.
The Biden Administration cannot mandate what states do. But CDC can change its guidance for states so that the communities most impacted by COVID are eligible to get in line now.