US vaccine rollout must solve challenges of equity and hesitancy
This essay is part of a PIIE series on Economic Policy for a Pandemic Age: How the World Must Prepare.
The effort to immunize the US population against COVID-19 as quickly as possible is one of the most complex logistical undertakings ever attempted in this country. Four obstacles have hobbled health authorities’ efforts thus far: The supply of doses from the manufacturers has been limited; bottlenecks have disrupted the pipeline from manufacturers to recipients; too many people are hesitant to accept the vaccines; and the distribution of doses across the population has been grossly inequitable. Recently, however, supply has been increasing, bottlenecks are being worked out, and hesitancy has receded.
But more effort must be invested to persuade people of the value of the vaccines and to reach all segments of society, particularly communities of color, immigrants, and those least able to bear the costs of the disease. These steps are all the more urgent because pandemics are on their way to becoming a chronic aspect of the US and global landscapes.
As might be expected with such a massive and complex effort, many aspects have been chaotic, frustrating, and inefficient. More planning went into the development and manufacture of vaccines than into their distribution. State health authorities reportedly failed to move aggressively in thinking through the complicated steps that would be involved in reaching all corners of their populations. They rightfully complained that the federal government provided shifting and late guidance on the number of doses they would be allotted. Hospitals and other elements of the health system already overwhelmed with caring for desperately ill patients struggled to divert personnel and other resources to the task of injecting shots into arms.
Initially, capacity for manufacturing the vaccine was limited. Even by the end of January, the federal government had delivered only 50 million doses to state and other similar authorities. Given that two doses were required of the vaccines that had been granted Emergency Use Authorization, that amount represented less than 8 percent of the doses required to immunize the US population. Over the second half of January, only about 1.3 million doses per day were being delivered by the Centers for Disease Control and Prevention (CDC). If that pace had been maintained, the states would not have received enough doses to immunize the entire US population until the end of April 2022.
The distribution effort gains traction
Since then, the supply outlook has improved. As of March 7, more than 116 million doses had been delivered to the states, and the average pace of shipments over the week ending March 7 picked up to 2.9 million doses a day. Moreover, a third vaccine (the one from Johnson & Johnson) has been granted Emergency Use Authorization. This vaccine has the considerable advantages of requiring only one injection rather than two and of being storable for up to three months at normal refrigerator temperatures and up to two years at essentially the temperature of a standard freezer. Although only a few million doses of the J&J vaccine were ready to be shipped when the Emergency Use Authorization was granted, J&J has committed to providing 100 million doses by the end of June—enough to vaccinate 30 percent of the population—and should meet that commitment even earlier under an agreement that will have Merck help make the J&J vaccine. On March 2, President Joseph R. Biden Jr. announced that the United States should have enough doses by the end of May for every adult in the country.
During the early weeks of the distribution effort, state and local health authorities could not keep pace even with the limited flow of doses from manufacturers; in addition, some states and localities may have continued withholding second doses even after the federal government gave up on that strategy. For whatever reason, fewer doses were being injected into people’s arms each day, on average, than were being shipped to the states. As a result, the backlog of doses that had been shipped but not injected increased rapidly. By the second week of January, this backlog had moved above 15 million doses (see figure 1). Since then, as states and localities have increased their capacity to administer the vaccines, and perhaps as they stopped holding back second doses, the backlog has mostly trended sideways. Some of the fluctuations in the size of the backlog during mid- and late February may have been due to the harsh winter weather that disrupted both shipments and injections for several days. During the first week of March, more than 2.1 million doses were administered on average per day—the fastest daily pace yet, but still not as fast as the stepped-up pace of delivery. As a result, the backlog moved above 25 million doses in the first week of March.
Performance in administering the vaccine has varied widely across the states. One metric of state-level performance is the share of the state’s population that has received at least one dose of vaccine. As can be seen in figure 2, as of March 7, the worst-performing six jurisdictions had all administered a first dose to less than 16 percent of their populations. By contrast, the best-performing five states all had administered a first dose to more than 23 percent of their populations. Across the country as a whole, nearly 18 percent of the population had received at least a first dose.
No compelling explanation has yet emerged for why some states have gotten vaccine into arms so much faster than others. William Galston and Elaine Kamarck examined a number of possible explanations, including that having a small population might help, or that having more vaccination sites per capita might be a plus, or that higher-income states might do better. None of these potential explanations does a very good job of explaining the differences in performance. States and localities have thus far been left to figure out the retail piece of this operation on their own, and it could be that the best explanation for differences in performance is that competence in state and local administration matters.
With the supply situation clearly moving in the right direction and the kinks slowly being worked out of the distribution system, one critically important remaining challenge is that the distribution of doses thus far has been deeply inequitable. Although Black and Hispanic persons have died at far higher rates relative to white persons, Black and Hispanic persons have been vaccinated at much lower rates relative to whites. If vaccine doses were administered in proportion to the incidence of hospitalization or death, the pattern of inoculation would be reversed, with whites receiving a much smaller proportion of total doses than either Blacks or Hispanics.
A wide variety of issues could explain the many gross inequities that are evident in data on vaccination rates. Lower-income individuals may not have the same ease of access to the internet (crucial for registering for a vaccination) as higher-income people or may be more concerned about data charges associated with filling out an online form. Immigrants may be concerned about providing personal identifying information if they are in this country in an undocumented status. Pharmacies and hospitals may be scarcer in areas where minority populations are concentrated, and options for transportation to vaccination sites may be limited or inconvenient and may intrinsically involve heightened exposure to other potential carriers of the disease. Anecdotal reports suggest that even when local authorities set up vaccination sites in areas easily accessible to high-vulnerability minority communities, privileged whites in some cases have travelled to those sites, snapped up available slots, and thereby effectively denied doses to some of those for whom they were intended.
One final issue clearly has played a role thus far in impeding the rollout of the vaccine and may come to play a larger role as the supply of doses begins to catch up with demand: Hesitancy. Roughly one-third of members of the military have refused to take the vaccine. In the first month of the rollout of the vaccine at skilled nursing facilities, three-fourths of residents were inoculated but only a little more than one-third of staff chose to take a jab.
A series of surveys conducted by the Kaiser Family Foundation has shown that overall hesitancy has receded somewhat. In February, Kaiser found that 44 percent of adults surveyed either would “wait and see” whether to get a vaccine, would do so “only if required,” or would “definitely not” get a vaccine. This share was down from 63 percent in December and 51 percent in January, but it remained high enough to potentially impede the overall effort to disrupt community transmission of the disease.
Hesitancy may have played a role in generating disparities in vaccination rates across races and ethnicities. In mid-February (the time of the most recent Kaiser survey), the share of respondents saying they would either “wait and see” whether to get a vaccine, would do so “only if required,” or would “definitely not” get the vaccine was 57 percent for Black adults, 48 percent for Hispanic adults, and 37 percent for white adults. Although these percentages have declined since December for all three groups, they remain high and the differences across groups remain wide. Hesitancy is difficult to measure, however, and different surveys have found different levels of it among different groups. A Civiqs survey from March 8 found vaccine hesitancy is higher among white Republicans than any other demographic group with little change over several months.
But there are, unfortunately, compelling reasons for some of that hesitancy among Blacks. Vanessa Northington Gamble argues that too much attention has been paid to the distrust of African Americans for the medical system and not enough to the failure of the medical system to earn their trust. “The relationship of the African American community with the medical and public health communities did not begin or end with the syphilis study. There has been hundreds of years of mistreatment of African Americans within the health care system.”
Whatever the underlying reasons, as the overall supply of vaccines continues to increase and the bulk of the willing population is vaccinated in the next few months, hesitancy among Blacks could become a key stumbling block on the road to inoculating enough people to decisively suppress the community spread of the virus and slow the development of new variants.
Lessons for the pandemic age
Given the probability that pandemics will become chronic in the United States and around the world, decisive steps must be taken to address all four of the issues that have hobbled the rollout of COVID-19 vaccines thus far in the United States.
- To ensure that adequate capacity exists for the manufacture of COVID-19-related vaccines over the longer term, the federal government may need to pay substantial sums of money to motivate manufacturers to mothball (but not to decommission) spare capacity. Maintenance of manufacturing capacity may be less of an issue (or not an issue at all) if COVID-19 remains a present threat, requiring regular revaccination.
- To ensure sufficient capacity for administering massive numbers of injections, the federal government should invest in dramatically expanding domestic public health infrastructure. This effort will need to be undertaken in collaboration with state and local governments.
- To reverse the grotesque inequities in the distribution of this and future vaccines, countless barriers will have to be dismantled. The need here is particularly urgent in light of the life-and-death consequences of the inequitable status quo.
- To overcome hesitancy, broad public trust will need to be restored in science, evidence, experts, governments, and other institutions. Key messages will have to be delivered in ways that are tailored to many different segments of the population. Most fundamentally, the government will need to demonstrate that it deserves the trust of its citizens—even those who, for generations, have been mistreated and had every reason to withdraw that trust.
This post was updated on March 10, 2021, with a clarification on the issue of vaccine hesitancy.
1. The backlog calculated here represents the difference between (a) the number of doses reported as having been delivered by the federal government to state and other similar authorities, and (b) the number of doses reported as having been injected into people’s arms. The backlog, calculated in this way, may differ from the number of doses that states have available for future administration due to a couple of considerations: First, to the extent that doses have been lost due to spoilage or other issues, the remaining available number of doses will be smaller than the number reported here. It is clear that some spoilage has occurred; for example, in Memphis, TN, the top health official resigned following revelations that 2,400 doses had been allowed to expired. Second, there may be discrepancies between the number of doses reported as having been distributed versus administered due to issues related to how many doses were contained in each vial of vaccine. At some point, the realization was made that if especially efficient syringes were used, six doses could be extracted from each vial rather than the originally advertised five. However, efficient syringes were not universally available. It is not apparent from public reporting whether the manufacturers have been claiming six doses per vial, and it is not apparent how many doses per vial have been extracted in practice. These data-related issues could be solved if states tallied directly the number of doses remaining in inventory and available for future administration, but no such data are publicly available.
2. According to a CDC analysis of data through January 30, 2021, Black persons had died at 1.9 times the rate of white persons after adjusting for differences in the age structure of the two populations. Similarly, Hispanic persons had died at 2.3 times the rate of white persons on an age-adjusted basis, and American Indians and Alaska Natives had died at 2.4 times the rate of white persons. By contrast, Asian persons had died at the same age-adjusted rate of white persons. For each of Blacks, Hispanics, and American Indians or Alaska Natives, hospitalization multiples were even higher than the death multiples. An analysis from the Kaiser Family Foundation (accessed March 7, 2021) found that, in the 36 jurisdictions (35 states plus the District of Columbia) that provide information on the race and ethnicity of people receiving a first dose of the vaccine, 13 percent of white persons (as of March 1, 2021) had received a first dose, compared with 7 percent of Black persons, 5 percent of Hispanic persons, and 11 percent of Asian persons. In other words, white persons had received first doses at nearly twice the rate of Black persons and more than twice the rate of Hispanic persons. In five states (Pennsylvania, North Dakota, Utah, Wisconsin, and Arizona) the vaccination rate for white persons was at least 2½ times the rate for Black persons. Among these 36 jurisdictions, only Alaska reported a higher vaccination rate for Black persons than for white persons. For Hispanic persons, the comparable figures are even worse: Ten states (Georgia, Pennsylvania, Colorado, North Carolina, South Carolina, Indiana, Nebraska, Wisconsin, Arizona, and Oklahoma) reported vaccination rates for white persons that were 3½ times or more the rate for Hispanic persons. No state reported a vaccination rate for Hispanics that was as high as the rate for whites.
3. Inequities in the availability of vaccine clearly continue even now. For example, the Miami Herald reported on February 25 that in Florida, only 5.6 percent of vaccinations administered had gone to Black people, even though they constitute 17 percent of the state’s population. Similarly, the Washington Post reported on March 4 that in Maryland, Hispanics have received only 3.7 percent of doses administered thus far despite constituting 10 percent of the state’s population, and Blacks have received only 16 percent of doses despite representing 31 percent of the population.
5. One indication that vaccine receptiveness may continue to improve is that Kaiser survey participants with closer connections to someone who had already been vaccinated were more likely to say they would get vaccinated as soon as possible. See figure 11 in KFF COVID-19 Vaccine Monitor: February 2021.
6. The mention of the syphilis study is a reference to the notorious Tuskegee study in which African American men were told they were being given free health care but were not and in fact were being observed for long-term effects of untreated syphilis.