Professor Nancy Krieger (Harvard, UCB Alum) along with Professors Mahasin Mujahid and Corinne Ridell (UCB) engage in conversations about the impact of racial discrimination, social class and place on the excess disease and death rates from COVID19 among African American and other communities of color. The session focuses on some of the thorny issues related to collecting and analyzing relevant social data on COVID19; and also on advancing a social justice agenda in addressing racial/ethnic disparities in disease rates. The conversation will be moderated by Professor Rachel Morello-Frosch.
This event was part of the School of Public Health's Dean's Speaker Series; the 400 Years of Resistance to Slavery and Injustice initiative speaker series, and was co-sponsored by the Berkeley Center for Social Medicine.
Michael Lu: Welcome to the Dean's Speakers Series, hosted by the UC Berkeley School of Public Health. My name is Michael Lu, I am the dean of the school, and today we're gonna have a conversation on structural racism, social justice, and COVID-19 with a distinguished Harvard professor and one of our most accomplished alums, Dr. Nancy Krieger. Nancy will be joined by three of our own amazing professors, Dr. Rachel Morello-Frosch, Dr. Mahasin Mujahid, and Dr. Corinne Riddell. Now, the last time we had a Dean's Speakers event was about three months ago, on February sixth, and the world has certainly changed a lot since then. At that time, there was 10 confirmed cases of COVID-19 in the United Sates, and the Diamond Princess was still looking for a place to dock in the San Francisco Bay. Today, there are more than 1.1 million cases in the United States, and nearly 65,000 deaths. And worldwide, there are now nearly 3.3 million cases, and 234,000 deaths. Now, I have to admit that when I signed up to be dean, last July, I never saw this coming. But I have to say that I have never been prouder to be the dean of this amazing community, to see how our faculty, students, staff and alums have risen to the occasion to provide public health leadership at this moment of great public health crisis. So as we speak right now, many of our researchers are working day and night at the leading edge of COVID19 research, from basic science research, to better understand the molecular mechanisms of SARS-COVID2, to rapid diagnostics and novel therapeutics looking for game-changing solutions to this crisis, to population-based studies on statistical modeling asymptomatic transmission, and now safe strategies for reopening the Berkeley campus, the Bay Area counties, our state and nations around the globe. And many of our students, staff, and faculty have also stepped up to serve our communities, especially the most vulnerable populations, by organizing PPE drives, blood drives, food drives, and many other relief efforts. And of course many of our alums are working right on the front line, in clinics and community health centers, hospitals and health systems. And in public health, leading local, national, and global COVID-19 response. But one thing I'm most proud of is that many in our community have been a voice for equity and justice in this crisis, which is so much core to our collective DNA. They've shined a spotlight on COVID-19 health disparities and the structural inequities and racism which are the root causes of these disparities. So, it is with the greatest pride that I'm welcoming all of you to this webinar, to kick off this conversation on structural racism, social justice, and COVID-19 with one of our most accomplished alums and three of our most distinguished professors. To introduce them, I will hand the mic off to Professor Rachel Morello-Frosch, who's not only one of the most amazing professors at UC Berkeley, with a joint appointment between the School of Public Health and the College of Natural Resources, but also someone who's worked on health equity and environmental justice, and epitomizes what our community is all about. So please join me in welcoming Dr. Rachel Morello-Frosch, who will moderate this webinar.
Rachel Morello-Frosch: Thank you, Michael. It's my pleasure to welcome you to today's panel session: Conversations with Nancy Krieger on Structural Racism, Social Justice, and COVID-19. Before we begin, I'd like to acknowledge that UC Berkeley sits on the territory of Huichin, the ancestral and unseated land of the Chochenyo Ohlone. Every member of the Berkeley community has, and continues to benefit from the use and occupation of this land since the institution's founding in 1868. This event is sponsored by the UC Berkeley School of Public Health as part of the Dean's Speakers Series and the Berkeley Center for Social Medicine. Today's panel is part of a series of events launched by the Othering and Belonging Institute at UC Berkeley, commemorating the beginning of slavery in the American colonies 400 years ago. This anniversary was marked by the 400 Years of African American History Commission Act, national legislation that was created in 2018 to commemorate the impact of slavery and laws enforcing racial discrimination had and continue to have on the United States, and the extraordinary sacrifices and contributions of Africans and African Americans to building our country. Along with campuses across the country, Berkeley has been engaged in a year-long commemoration of this anniversary, and we have had previous events on the impact of slavery, Jim Crow, and how legacies of structural racism affect present-day community health and drive health disparities. Such panels have included discussions on the origins of American gynecology through experimentation on enslaved black women and the ongoing, lingering effects of this legacy on the health of African American women. Similarly, the legacy of Jim Crow, red-lining, and exclusionary immigration policies have historically shaped the disparate impact of epidemics and chronic disease on African Americans and other marginalized groups, and as we heard from last week's panel, Straight Talk: Structural Racism and COVID-19, this was a great conversation last week, and we hope to continue this chain on this panel. Recordings and transcripts of these previous panels can be found on the Othering and Belonging Institute's website. Today our speakers will discuss the legacy of slavery and Jim Crow as they unpack the role of structural racism in understanding the social determinants of COVID-19 and its disproportionate impact on African Americans and other minoritized groups in the US. Indeed, while the media and many public health officials tell us that viruses do not discriminate, we know that people and institutions do. The question before us, then, is: how do we systematically address this challenge as researchers, public health practitioners and social justice advocates in an era of COVID-19? To engage us in this conversation today, we are joined by Dr. Nancy Krieger and an extraordinary panel of experts who will focus on this issue. Dr. Krieger is professor of social epidemiology and American Cancer Society clinical research professor in the Department of Social and Behavioral Sciences at the Harvard T.H. Chan School of Public Health. She's an alumnus of UC Berkeley, having received her PHD in epidemiology in 1989. She's an internationally recognized social epidemiologist, with a background also in philosophy of science and the history of public health, combined with over 30 years of activism, linking issues involving social justice, science, and health. She is also co-founder and chair of the Spirit of 1848 Caucus of the American Public Health Association. Dr. Mahasin Mujahid is associate professor and Chancellor's Professor of Public Health at UC Berkeley in the School of Public Health, in the Division of Epidemiology. She is a nationally recognized social epidemiologist whose research examines neighborhood health effects and links to racial/ethnic health inequities with a focus on cardiovascular health over the life course. And Dr. Corinne Riddell is assistant professor of biostatistics at the School of Public Health. And she's a perinatal and social epidemiologist whose work examines structural links to racial inequalities in infant mortality, as well as mortality due to other causes. And she also examines why some states have fared better at reducing health inequalities than other states. So the format for today's panel will include conversations by our distinguished guests, followed by a question and answer period from our audience on Facebook. So, starting at the present moment, the media has been covering the disparate impact of COVID-19 on communities of color, and yet there's been little discussion of the role of structural racism in shaping how this pandemic is unfolding in the US. So Dr. Krieger, you and your colleagues released a paper last week that examines US and county-level and zip code characteristics that can inform strategies for developing more equitable COVID ID response. Can you tell us a little bit about this analysis and what you found?
Nancy Krieger: Sure, first I'd like to say thank you very much for including me in this panel. I'm really heartened to see that there have been a variety of panels happening around the country, I've seen them posted on 1848, that are really trying to get at how racism and structural racism, in particular, and the histories of this country with regard to histories of enslavement, with regard to histories of colonization, with regard to histories of borders and contested borders, are really shaping the current contours of this pandemic, as it's playing out in the United States, both in terms of who's infected, and also what the policy responses are. So it's really great to be part of this particular one, as a discussion, as well as being a growing community of voices, individuals, organizations, and more, that are coming together to start to articulate this framework, because it's really crucial. Because there's so much work that goes on, that continues to be victim blaming, and it's really critical to have a structural frame to begin to change that. What we did in our analysis is that noting that there were, for example, very few data being reported, and this is still a major problem, particularly for cases and hospitalizations, by race/ethnicity, or any socioeconomic measures whatsoever, noting that even the death data still are deficient on what's being reported with regard to race/ethnicity, although that's getting slightly better, but only from some states, in terms of the amount missing. We took a strategy which we've used now for a long time for our public health disparities geocoding project, saying: well, wait, it's a death record, or we know something about where it's happening. We can look at the characteristics of counties, we can look at the characteristics of zip codes and we can see: how do rates differ by differing groups, whether it's a county or a zip code that has high poverty, or low poverty, whether it has a high concentration of household crowding, or a low concentration? And that's, and whether it has a high concentration of people of color or particularly, say, for example, black or African American versus a low concentration. And that's what we did. And what we were able to document in the first data showing this, because the point is: sometimes you need data to show what people actually know, what the journalists have been doing an amazing job of reporting, because you actually need the numbers and you need the evidence both to sway policy, to marshal resources, and to also point to where intervention should be put in terms of whether those interventions are, for example, point and place, crop up crowd, pop-up testing centers, whether it's where there are communities where personal protective equipment ought be distributed by the local health officials, et cetera. And so there's some key findings, that we found stark social gradients. So what we did was also, rather than just simply comparing the number of cases in one place versus another, we actually looked at the rates per 100,000, and just for example in the case of looking at the counties by the COVID death rate per 100,000, was that in counties with very high crowding, their death rate was nearly 17 per 100,000, and it was under five for those that are counties with low crowding, and it's really important to know that the census variable that exists in crowding, it's basically you have to have more than one person per room, not counting the bathroom, the hallways, and the kitchen. And what that means is that if you have a one-bedroom apartment, sorry, no, you do get to count the kitchen, but you don't get to count the bathroom. So, if you have a one-bedroom apartment, which has one bedroom, one sitting room, one dining room, and one kitchen, it's not crowded until there are five people living in it, by definition, so it's really important to understand that it's restricted to definition, just as the poverty cut point that's used in this country, if you are under the poverty cut point, you are seriously impoverished, and that's why many health programs even use a higher cut point and go up to 200 percent of the poverty line. So, we found that repeatedly for the poverty line by populations of color that there's higher rates of the death, and higher rates of what is being reported for the testing on positive tests, but those are very suspect data, because they are no population-based; they're sporadic. You don't know who's actually getting tested. So it's difficult to make inferences. The death data are more complete. So it's striking gradients, and it's bringing the socioeconomic element into understanding the racial/ethnic inequities that are also being reported, even as the data on race/ethnicity continue to be abysmal with regards to their missing-ness.
Rachel Morello-Frosch: So, you looked at a variety of social determinants in relationship to this data, so can you say a little bit more about what the implications of this work is for future data needs, in terms of COVID-19 surveillance strategies and its implications for how we try and address this pandemic?
Nancy Krieger: Right, so one thing to be very conscious of is that in being concerned about the state of the data, it has to be framed at the time of understanding there has been considerable un-investment and lowering investment in public health for decades in this country. So health departments, local and state, are overwhelmed and underfunded and understaffed. So one has to give credit to people trying to get the best data out they can, and the public has stepped up in all kinds of crowdsourcing ways to try to improve that, and the CDC has been diminished in what it can do, and that's completely separate from whether it's being politically curbed in what it can do. So it's important to be fair and understand at the same time, there has been attention in this country to preparedness for many different kinds of disasters and emergencies, and from that standpoint, it would seem that there should have been a data plan in place that makes it clear that key data are for the localities to understand what's going on. And localities understand what's going on with relation to their racial/ethnic and economic composition. That's huge, as well as, of course, age and sex/gender. So the opacity of the data are really alarming. What these data point to is the need to, one, improve the data that are being reported at the individual level so that there are individual level data. For example, the death certificate routinely includes data on race/ethnicity, on education, on occupation. The occupation data may not be so great, in terms of the quality, but the education is usually pretty good. Not to be reporting these data is, is just, makes no sense. So, it makes one wonder why it's not being reported. But it should be there. And the other data, it makes no sense whatsoever that when people are being tested that there cannot be, for example, a simple little digital form, not a paper form which is part of the problem, was the CDC had developed a form that was paper and took over 30 minutes to fill out, that obtained demographic data and lots of other detailed data, and nobody filled it out. It was impractical. To have a simple, digital form that has the exact same sociodemographic variables as appears, for example, on the death certificate, which also, by the way, in addition, has the zip code, so you can also link to zip code. Zip code is not great; census track would be better. But great, not great is a lot better than nothing at all.
Rachel Morello-Frosch: Okay, so kind of feeding on that answer, I'm wondering if we could turn to Mahasin, and then if Corinne can weigh in as well: why is it so important to integrate place-level data in our analysis and surveillance strategies, and in particular in the case for COVID-19, and how do we better integrate individual and place-based factors into these surveillance efforts? So maybe Mahasin could start for us.
Mahasin Mujahid: Sure, I mean, I think obviously as a social epidemiologist who studies the health impacts of place, I understand the importance of these kind of data, and why it's so critical to have analysis on this kind of data, but I think I just want to sort of emphasize two things, two important points, for those of you that are less familiar. First, the idea that we need to have information on place and describe health by place is not new. In fact, we've been doing it for some time. It's one of the hallmark features of what we do as epidemiologists, where we describe disease based on person, place, and time. And so we have powerful maps on major chronic diseases, mortality rates that vary across the state, across counties, across zip codes in the United States, and we documented important regional differences. As an example, we show that, you know, there are high rates of mortality in the southeastern portion of the United States, we've documented urban/rural differences, so in a lot of ways, data that we're now seeing in relation to COVID-19, specific mortality and it being highest in counties with higher levels of poverty or worse inequality is something that's consistent with these patterns that we see for a wide range of other health outcomes. I think the challenge is to unpack the reasons why we have this spatial variation, and we need to frame it as more than just the clustering of high-risk individuals on a particular location on a map. And I think to do this, we can tap into what is now a very rich literature that is highlighting the important ways in which the differential distribution of resources and opportunities, as well as stressors across places, become embodied and lead to disparate health outcomes. And so I think having these data are a critical first step in terms of understanding the transmission of COVID-19, and also understanding which communities are more vulnerable to the potential impacts of COVID-19. So there are some similar kinds of data that are coming out now. One example of that is from the Brookings Institute, and they released data showing that using an equity index that they developed, and that considers the poverty in a county, as well as racial gaps in life expectancy, and multi-generational family cohabitation, sort of in an area. And they're showing that the places that are hardest hit by COVID-19 are the same places that are high on this equity index, and those are places like Detroit and New Orleans, as an example. And then there are data coming out of Harvard University that show that COVID-19 related deaths are highest in counties with higher levels of air pollution. So having these data allow us to understand what is putting people at risk in these areas, not only of contracting the disease, but also in terms of the long-term consequences of the disease. And these consequences we have to understand as more than just the health-related consequences, but also the social and economic consequences that can further exacerbate the kinds of health disparities that we see.
Rachel Morello-Frosch: Thanks Mahasin. Corinne, I don't, ah, do you have anything to add about this issue, about how we better integrate individual and place-based factors?
Corinne Ridell: I think Mahasin did a really good job of covering kind of the spectrum of things that we want to consider, but just to echo what she was saying is that, when we look at place, we can then link people by where they live, and then kind of link onto the data, looking at policies or these area-level measures or individual-level measures, and then we can see what relationships exist in the data, like Mahasin was saying, and that's really important. And allows us to then figure out ways to intervene.
Rachel Morello-Frosch: Great, thank you. So this leads me to my next question, which is: how does trying to more get at this question of structural racism and the way it drives health disparities, in this case, disparities in the impact of COVID-19, how does that framework help us think about how otherisms play out in the COVID-19 epidemic? What kind of intersectional thinking should we science and public health practitioners be doing to better understand the elements of gender or class, heteronormativity, et cetera, in the work that we are doing to address and understand the drivers of this pandemic, and how we intervene to protect communities? So maybe Corinne, I'll start with you.
Corinne Ridell: Sure, thank you Rachel. So just to define for those watching, structural racism is a set of systems and structures in place that distribute resources and thereby wealth and health, disproportionately across races and ethnicities, away from people of color. And we're seeing this manifest through the disparities in mortality from SARS-COV2, with higher mortality rates among the black community. And we know there are many -isms, such as sexism, ageism, ableism, and I'm gonna focus on -isms faced by the LGBTQ community. So there are homophobic and transphobic structures and systems which are associated in pre-COVID times with the LGBTQ individuals having more precarious home and housing situations, and their being at a higher risk of intimate partner violence, sexual violence and suicide, compared to other, non-LGBTQ people. And so that sheltering in place, while it's our best mechanism right now to stop the transmission of the virus, so it might protect them from the virus, it may also put these individuals at increased risk of violence and mental health distress. And then on top of this, others in that community can't shelter in place because LGBTQ youth are more likely to be unstably housed. And then in terms of your second question, on intersectional thinking, one thing that comes to mind, and I think it was mentioned during last Friday's panel, it was about home healthcare workers. Nine out of 10 of these are women, and half are black and Latin-X, and 15 percent are immigrants, and 13 percent are not US citizens. More than half of these individuals have a high school education or less, and they make about $13,000 a year or about $12 per hour on a part-time basis. So this particular workforce is composed primarily of persons of color who are women with low earnings, and I think what this shows us is when we are looking at the data, we have to dissect it in many directions, and we do that simultaneously, and what that allows us to do is identify individuals who may be at high risk because of increased exposure to the virus, or indirectly through an impending recession. And with that, we can make specific considerations for these groups who are at amplified risk and address these accordingly. If we have any hope of addressing the disparities that are resulting.
Rachel Morello-Frosch: That's great, thank you. Nancy, what are your thoughts about this issue? Otherisms?
Nancy Krieger: So the thing is that we live in our bodies as unitary beings who are subjected to multiple systems simultaneously, so it's not like one day you are a woman, man or you are transgender, it's not like another day you are white, black, Latin-X, Asian, American Indian, with all the kinds of categories that the US census comes up with; it's not like you are one day working class and another day professional. You're all these things at once, and so what is important about understanding this epidemic right now is that there's two different things going on. One is: what's the state of people and their health? Their physical and their mental health at the time this epidemic hit, and then how is that intersecting with both exposure and then susceptibility and risk of increased mortality? And so that really matters, because what's happening now is even if you, for example, got rid of all the inequality, but you left everybody in the same bodies that they're in right now, because of past inequality, that becomes embodied. And so your body is, all the time, doing this integration that intellectually one is asked to do with, quote/unquote: intersectional approaches. Because the body is not saying: today I'm this, or I'm that, and remember to put these things together. It's doing it all the time. So from that standpoint, you have a better understanding of who is being made vulnerable just because of their, the past will come into this more, the past experiences of injustice that already have led to incredible health inequities that occur in this country along these different dimension. And at the same time, it's also affecting the risk of exposure, the likelihood of being vulnerable, to greater risk of bad outcomes, and increased risk of death. So it's, for me, it's impossible not to think, quote/unquote, intersectionally, because I think about embodiment and I think about the integration of these exposures and experiences, and that's what translates to what leads to the increased populations rates of health. And what's really important is to think about this: we're individuals, we're always simultaneously, they're quirky selves is being whoever they happen to be in the world, and also shaped and shaping the world around them by virtue of the institutions that exist. And it's also not that these institutions are doing things passively. Sometimes there is the impact of decisions long ago in terms of, for example, who is allowed to have Social Security versus not when Social Security legislation was first passed, and domestic workers and farm laborers were cut out of it, and then that continues. And that's a legacy, not currently enforced, but it was enforced and had intergenerational consequences. But there's also very active attention to the question of: who is getting resources now because of denying it to others? And that always has to be part of the equation. Watching the discussion happen, quite fascinatingly, between Mitch McConnell and Governor Cuomo was quite instructive, where when as Mitch McConnell was suggesting that New York state should not be given additional resources, Cuomo was pointing out, actually, states like his and yours like California contribute more back to the federal budget in terms of what you pay than what you get. And Kentucky actually gets a lot of resources. So it's really important not to only be looking at who is being, quote/unquote, deprived and denied, but who's gaining from this. And that is also, has to be a piece of trying to say that you're having a structural analysis, because it allows you to look at what the policies are that need to be changed to actually deal with the injustice.
Rachel Morello-Frosch: Great, thank you. So, putting this current work on COVID-19 and structural racism into a broader context, Nancy, you've done a lot of fabulous prior work, looking at how legacies of Jim Crow and slavery continue to shape health disparities today. So can you talk a little bit about that? You know, what is the connection between the history of Jim Crow and slavery to COVID-19. Will you talk a little bit about some of your prior work on this question?
Nancy Krieger: So one thing I want to just say right now, it's still empirically unknown, in terms of looking at Jim Crow and COVID-19. You can make assumptions as Mahasin was already pointing out in terms of the concentration of both the worst mortality, predictably, among black populations, albeit with inadequate data on the death certificates, for perhaps southern states. So there are ways, but what you can do is you can first step back and say: what has research already shown about Jim Crow? And Jim Crow obviously comes right on the heels of slavery, in the sense that there was Reconstruction after the Civil War, there were some of the first major efforts at having not only greater enfranchisement, but civil liberties, civil rights, economic development, and then there was a severe backlash, and that's what the undermining of Reconstruction was about. Part of the undermining of Reconstruction was the development of the Jim Crow laws that effectively enshrined white privilege as well as denied privilege to others. Predominantly black, but not only. It also applied to people who were Mexican American, American Indians and others, and it was, these are important legacies and an important thing to remember is that it's often treated like that was a very long time ago. But Jim Crow didn't end until 1965, and the act prohibiting discrimination in housing is not until 1968. 1965, someone born in 1965 is now 55 years old. So that's really important, because it means that anyone older than that could have potentially been born in a Jim Crow state, under Jim Crow conditions, and early life exposures really matter for subsequent health. But also it means that, when were your parents born? And where were they, and what were they exposed to? That's gonna affect people that are less than, you know, that are more than 55, et cetera, et cetera. So you start thinking about the fact that Jim Crow can leave an impact now; we've done work showing, for example, what the, that the end of Jim Crow led to a marked reduction in the black disparities in infant mortality, comparing people in Jim Crow versus non-Jim Crow states. And that tells you something about, those are the people born 55 years ago, which is really big, because those who survived are the ones living now, and are in this group, an age group that are particularly susceptible around COVID-19. But we've also done work, for example, I've led work looking at how the type of estrogen receptor for women who have breast cancer among black women depends, in part, women now, living now, on their Jim Crow birthplace, and when they were born, because again, breast cancer is, one of the conditions, it's a cancer, having cancer, if you have cancer and you have COVID-19, that's not a good combination. So, those are ways you can imagine it, literally, right now, affecting, as I was saying before, the state of people's health, and their bodies, that would be directly, potentially related to increased risk. But there's also the economic reasons, and there's studies that have looked at the intensity of where the slave plantations were biggest and worst in relation to where they're located in Jim Crow states, in relation to current levels of economic impoverishment. And seeing that there's generations of inequity that get put in that way, does that also affect what's going on with COVID-19? I mean, those are reasonable questions to ask. But I'd also wanna flag, there's another part around histories of enslavement in this country that aren't discussed as much, which is that American Indians were also enslaved. And California certainly has its history with regard to enslavement, and so that history has to be brought in in terms of, also, where does questions of colonization fit in? Because obviously the enslavement of Africans in what's now the continental US depended on there being colonization of the US, as well. And so that intersects, and right now there's huge concerns as people know about the growing toll of COVID-19 in Navajo Nation, concerns of what's happening in Oklahoma. Oklahoma also actually had Jim Crow laws. So these connections come together that are really important to think about in terms of how the past continues to feed the present, by shaping people's bodies over generations.
Rachel Morello-Frosch: Thank you. Mahasin, what are your thoughts on this, in terms of how we understand the history of Jim Crow and slavery for current community health, and in particular, how we might think about it for understanding COVID-19, even though the data's just emerging, as Nancy said?
Mahasin Mujahid: Sure, so I don't have much to add. I think Nancy did a great job, but I will kind of highlight two points. The first is that I think, in our profession, we fall into these habits, not just in, ah, sort of epidemiology, but in other professions, where we use certain acronyms and certain jargon, and catchphrases, and one of the ones that we like to use that's really popular right now is this idea of really looking at the root causes of disease. And so we think about factors related to neighborhood conditions, we think about socioeconomic factors, we think about particular psychosocial stressors, and how those things might be root causes of disease. And I think we have an opportunity right now, it's a critical time for us to do so, is to be more intentional about these kinds of catchphrases. So if we're talking about root causes, then perhaps we need to deepen the conversation and think about, well, what soil do those roots exist in? Or another way we could phrase it is: maybe we need to trace those roots even deeper. Whichever way, whichever tagline will get us an explicit acknowledgement that what we are seeing play out in terms of, across racial and ethnic lines, across socioeconomic lines, is a function of the legacy of slavery, of colonialism, of genocide of native people, and we have to really be clear on that. I think my second thought is that, as epidemiologists, we often think about what we call counterfactuals, and so it's this idea that if we had a perfect world, we would follow a group of people who were exposed. Let's say they smoked cigarettes, or they'd been exposed to some kind of cigarette exposure, and then we followed them over time to look at their outcomes, for example, lung cancer. And then in an ideal world, you'd go back in time and make those same individuals unexposed, and then follow them over time to see how their outcomes would change. And so I would be remiss if we didn't talk about potential counterfactuals of two things that are actually receiving a lot of attention right now in the media. One is that Georgia is receiving a lot of attention, and it's based on this idea of removing their shelter in place and the unique timing of that decision, with the MMWR that just came out from the CDC, that documented that over 80 percent of the hospitalizations related to COVID were in fact in blacks or African Americans, and so the counterfactual question would be, well, what if that 80 percent were white? Or over 80 percent were white? In a different vein, but also important, in the media we are seeing information about the protests that are going on in the state of Michigan, so we know that, for example, there were armed protesters that went into the state capital building. Another important counterfactual: what would have happened if those protesters were black? And you know, the thing about counterfactuals is that we can't actually observe the counterfactual, so as epidemiologists, we've come up with all of these powerful tools, a range of very rigorous methods to approximate those counterfactuals. And I think, unfortunately, we need to dump out our tool kits, like our toolboxes, and start over completely, from scratch. And we need to begin with putting into our toolbox a better understanding of how these historical insults are playing out, time and time again, and then we can add that other suite of rigorous methods, but we have to really be clear on starting with these historical insults and understanding that if we don't, if we don't acknowledge how they're playing out, then we actually can't achieve the health equity that we have set for ourselves as a national agenda.
Rachel Morello-Frosch: So that gets me to my next question, because what you're saying sort of compels us as scientists and epidemiologists, to really do more to lift up this connection between structural racism and health. And yet, we get a lot of pushback, there's lot of pushback right now, on folks that are, you know, being, having public conversations about the very issues that we're discussing right now. So as scientists and as public health practitioners, how do we do more to lift up this connection in our research, and I think even more importantly, how do we disseminate the implications of this work that we're doing for understanding these structural causes to broader audiences, including policymakers and decision-makers, who are really sort of making the decisions that have life or death consequences for communities of color? So maybe Nancy, you wanna start?
Nancy Krieger: Sure, I mean, I think one very important place to start is to say the reasons for asking these kinds of questions is not to be politically correct. It's actually to do rigorous science, it's to do correct science, because by not asking about these important determinants, you miss part of the story, as in, a huge part, and you don't understand what's going on and why risk is what it is. If you end up sticking with individual level, victim-blaming behavioral models, then you end up lecturing people about behaving better, and you forget the fact that a third of the US is at risk of not having their water on, because of problems with payments to utilities. So it's a little hard to tell people to wash their hands if they don't have running water. So this is, it's critical, and that's specific, by the way, and leads to many different things that are sanitation-related, but to COVID-19. Because another thing that's really important is, there's a lot of tendency right now to sort of act, use COVID-19 to highlight something that you might have cared about already, beforehand, but I think there's a different way to understand what COVID 19 is doing, which is that, one, you're showing us simultaneously our common humanity, in terms of what we can be vulnerable, as a species, to this virus, and secondly, showing us what our profound divisions are. It's not either/or, it's both/and. And we're learning new things every day, the we in that sentence being scientists and journalists, because they're uncovering tons, sometimes before the scientists are, about what's actually going on. There's still so many questions. There's questions about what's going to be classified as a COVID death, because of more organ systems are being affected; there are so many different things, what's going on by different age groups, that are unknown, so there's going to be a lot that's truly new knowledge, that's novel. No one has seen this before. No one knows what the answers are going to be about which antibodies you have exactly when, for how long, if you're going to be immune or not. There's many, many uncertainties, and one thing I think, also, is that people are used to wanting to deal with risk, meaning you know what the parameters are and you can estimate what the risk is, and here we're in a different world, which is uncertainty. There's so much not known, and the future is not known, and there's so much that I think we're confronting with that so I think another role as a scientist is to say: therefore, we need to do what we can to be informed by the best science possible. And the best science possible, in the case of COVID-19, requires dealing with health inequities, because that's exactly how this pandemic is playing out.
Rachel Morello-Frosch: Okay, so, just a little bit of followup on that. I think in public health we talk a lot about health inequities, and then that's, I think, an easier conversation. But I think where there's a lot of reluctance and trepidation is conversations around racism, structural racism and why it's important, scientifically, to have this conversation amongst scientists. So, I'm, I feel this in my own work, as an environmental justice researcher. I talk, use the words: environmental racism. So I don't know if you have anything to say about that, but.
Nancy Krieger: Well, again, if part of the thing on structure, just to jump in on structure, is to then point that if there's structural problems, that means structural solutions, and it puts things towards the policy realm, and the realm of what advocates are often pushing for and the relationship between that, so that if you don't do this, you're not going to, perhaps, change things that could be changed. I think what COVID-19 is doing is exposing, for example, back to the example that was brought up of home health aid workers, one, what kinds of benefits do they get? None, actually, so how do they benefits? How can they be organized into, what kind of unions are feasible? What's happening around sick leave, paid family leave? But you it by drawing in this history: why do certain people not have the, these kind of aspects to their job as part of what is expected for their job? And we have, it's infuriating to hear all this homage to essential workers who are not treated as essential when it comes to their personal protective equipment, or to what their pay and benefits are, and their ability to take sick leave. So, a structural approach to understanding these problems, and understanding where the structural racism comes in, is both about the policy change and it's also about recognizing where there are going to be communities in greater need, and they're in greater need because of historical injustice, and there's no way to get a shortcut around that.
Rachel Morello-Frosch: Yeah. So, Corinne, I'm gonna turn to you, but we are, you guys are distinguished panelists of social epidemiologists, so how can social epi elucidate solutions for addressing the COVID-19 pandemic, going forward? And why are theory and data and empirical research important for identifying solutions? So, Corinne and then Nancy would be great.
Corinne Ridell: Sure, so the way I think about this is I think about the data first, and that's probably my own bias, because I'm trained originally as a statistician. So I think about the data, and we know they're important, because if we can't measure the disparities, then we don't even know they exist, and we don't know to what magnitude they exist, or for which groups, so the data are really important for that, but data by itself isn't enough. And we really need theory to help us understand the data, because without this theory, then I think what happens is we stop with explaining away the differences we observe, and this gets back to what Nancy was saying, the victim blaming, so what we do is, what some people do, is that they say: oh, they're due to underlying health differences or differences in income. And what that can do is just alleviate a response, because it places the blame on the individual. But the theory really asks us to go deeper. So I think it's really echoing the last responses. So the theory asks: why are there differences in living conditions, why are there differences in health conditions, why are there differences in who composes the essential workforce, and who receives little pay? So that's what theory brings to the table. And then in terms of empirical research, the empirical research, I see it as really bringing together the theory with the data, and it allows us to describe, do all the things epidemiologists do, which is to describe trends, really important right now to conduct these surveillance studies, it allows us to conduct, investigate causal questions and to predict our healthcare needs, and in particular, I think it's, this is really important for causal imprints, because theory provides context for the disparities that social epidemiologists estimate. And it informs how we build our models, and so like Nancy was saying, it makes us build better models, because otherwise, you're leaving out part of the picture. And I do think that social epidemiologists are best-equipped to answer these questions because they're up to the challenge of using observational data to describe disparities and to investigate causal questions using quantitative methods, so they have, like, that quantitative toolbox. But now they're pairing it with theory. So I think social epidemiologists are the people that can do that.
Rachel Morello-Frosch: Thanks, Corinne. So Nancy, thoughts about why theory, data, and empirical research are important for solutions?
Nancy Krieger: So in my own work, mainly I focus on critiquing and understanding and developing epidemiologic theories of disease distribution, methods to test ideas, and also do the empirical, substantive research, and I think it's really important not to put theory in one unitary box. I mean, theories are about sets of interrelate ideas that scientists come up with if they're talking about scientific theories, there are obviously theories in literature, art, and many other fields of human inquiry and endeavor and appreciation, that are trying to explain phenomena, and explain them in order, in epidemiology, to also influence them. And so there are different kinds of theories. The dominant theories are lifestyle and biomedical in this country, they're used by most epidemiologists. Social epidemiology is a subset. Among social epidemiology, you'll find a range of theories, psychosocial, social determinants of health, there's the eco-social theory of disease distribution that I've worked on; there are others, and they take on different pieces of this picture. So there's not just one theory, but the idea is that you actually have theoretical frameworks that actually guide your understanding of what, causally, is going on. You don't get causal inference from the data itself. You actually have to have, you know, and it's more than just a model, because theories are much more encompassing than what a model is; a model is just a depiction of a particular set of hypotheses. So you have to start with theoretical grounding, and it will tell you what the data are. Data, as a word, means: that which is given. But it's not given, it's produced. And it's produced by people, and the question is: who's at the table, producing those data? With what resources? So that we do see that there are problems, as opposed to: no data, no problem. So for me, the sequence is always theory, to what it means around resources to get data, to get those data, and then to start testing ideas. And the other part that's really critical is that we do test our ideas. They're not just meant to be opinions, or personal revelations. So, I think that those become the relationships, but I also think that epidemiologists, as such, have a particular role to play. They've had an enormous role to play in the modeling, it's exposing what some of the different assumptions are, because it's theories that help you see the assumptions. So there are very different assumptions behind models that are based on biological principles and that are based just simply on crunching huge amounts of data, and they're behaving very differently, and that's a longer discussion. And so, but the, but we can do that, we can do the descriptive work, which is utterly essential, because that's about getting resources to where resources are needed now and understanding what need is. And that's crucial, and then yes, there is gonna be testing of ideas to explain what some of those distributions are, but that's just one piece of the evidence. Because there are policymakers, there are advocates on the ground, there are people that are doing mutual aid societies, so many different things. We're not creating the solutions. We're showing what the problems can be, and thinking as people try to make inferences based on incredibly biased, non-representative samples of people who managed to get tested, versus who hasn't been tested, we get to point to what problems are of existing data, what data need to get filled in, to understand what the real scope is of this pandemic, in the US and globally. And we get to help inform the discussions with evidence, but we are not, ourselves, the purveyor of solutions. We work with others; that's a social project. And that involves all of society.
Rachel Morello-Frosch: Great. Mahasin, your thoughts on this question?
Mahasin Mujahid: I think I would just like to give a caution, that the points that both Nancy and Corinne emphasized on the need of really unpacking the why, having the theory inform the models, is really critical in this time of big data and data science. We have massive amount of data, and the ability to click a button and have something input, and something sort of spit out. And then we run with that, and I think, you know, there's a lot of concern about whether or not big data will lead to misinformation, and will do more to reinforce stereotypes about marginalized populations. And will lead to more victim blaming and shaming. So I think now more than ever, we really have to have strong conceptual and theoretical understandings of the problems that we want to study, so that we don't take advantage of just the convenience of massive amounts of data.
Rachel Morello-Frosch: Great. So, are there contemporary examples or historical lessons that we can learn in terms of the role of epidemiology in elucidating theories of disease causation, that can move research in the direction of these structural issues that we've been talking about? So, yeah, Corinne, any examples of historical lessons that can inform us in terms of how we move forward with the field?
Corinne Ridell: Sure; one historical example that I think many of us are taught as a very first example in bioethics is the Tuskegee syphilis study, which was this decades-long passive surveillance study of the natural progression of syphilis, right, among black men in Alabama that ended in the 1970s. And econometric analysis has linked this study, in particular, with medical mistrust and decreased life expectancy among blacks, especially among black men. That's just one example of the way seeds are set in the past and lead to medical mistrusts still now, and how this like historical experimentation on black individuals in the United States, and so when we think about how we're going to respond to the pandemic in a situation where there's medical mistrust, there's been studies done that have shown like, if you link by race a person with their healthcare provider, they're more likely to have healthcare utilization and follow recommendations. And other studies that have shown, in the black community, that if you use barbers to measure blood pressure, and then have those barbers connect people with high blood pressure to pharmacists, so kind of taking out the medical care provider, then the group that got that treatment will actually have much lower blood pressure. This is in a randomized trial, and so for the pandemic, I think what we have to do is think about how we can develop innovative strategies for testing and contact rationing, that are adapted to reach hard-to-reach communities. And these communities are the ones that overlapped with being the hardest hit. And so that, our strategies will benefit from using the trusted institutions like the barber shop, rather than bringing in unknown outsiders into communities.
Rachel Morello-Frosch: Okay, so now I want to turn, thanks Corinne, I think this is a great way to end the conversation and get to audience questions. So, the first question I wanted to ask was: how can we ensure Native American data is actually reported and not just stuck in the quote/unquote "other" box? Corinne, your head is nodding.
Nancy Krieger: I can speak to that, also.
Rachel Morello-Frosch: Go ahead, go ahead, Nancy.
Nancy Krieger: My colleague Randal Aqui, who's he's a native Hawaiian down in UCLA and he's just got a paper accepted, and we had been in touch about it before, looking at the Indian Health Service data, to look at people who are on reservations, to get some of the data out there about what the inequities are. The problem is, of course, as you know well in California, that there's high concentrations of people who are, identify as American Indian, who are not part, on the reservation, and they're not counted well. So the American Indian epidemiologic groups are trying to figure out exactly how to get some of those data, but the problem is there, there's two problems. There's one getting the actual case data, and then here's also the problems of the denominators, how well they're counted or not. So, it's a, that's an ongoing issue, but there's definitely work and studies afoot to do that. And the other thing is, at the same time, is that, like, for data we're looking at in Massachusetts, if some populations are small, you describe what they are, you may not be able to do statistical tests, because small groups are small groups, but you don't hide the data in other. So you also have to be true to the data, because if it's, again, a very little group, it's going to be little and that has sample size considerations, but then you absolutely always publish at least the descriptive statistics, even if you can't necessarily do some of the more refined statistical testing.
Rachel Morello-Frosch: Right.
Corinne Ridell: Can I add to that quickly?
Rachel Morello-Frosch: Yes, go for it. Go for it, please.
Corinne Ridell: Like, in terms of death certificates, we know that even pre-COVID deaths to this group are 30 percent under-counted, and that's because of this mismatch between the numerator and the denominator. The numerator being the actual deaths, and the denominator, who is in the population. So it's the death certificates aren't being indicated as being the death of a Native American person, and so that's because who fills out that certificate isn't identifying the person, the decedent, correctly. And so I think in the, to measure this correctly, it's not just like small numbers because they're a small group. It's small numbers because they're being misidentified on the death certificate, and so if we could, like, tackle that and get, you know, race and ethnicity better identified on the death certificate, I think that would help us measure the actual size of the impact, because otherwise, they'll still be under-counted on the death certificate.
Nancy Krieger: And when the under-count is actually more like 50 percent, and it's also where our data linkage projects that are underway that are linking death certificate to other health records to also Social Security information into other kinds of things, that there's a really important, and that needs to be funded, to make that kind of work fly so that there can be actually better estimates.
Rachel Morello-Frosch: Yeah, this gets back to our public health and data infrastructure funding challenge, I think, as well. The next question, ah, is: have there been any comparisons yet between areas with and without intervention? So, efforts for example like bans on utility shutoffs or eviction moratoriums, things like that? Are there efforts underway to try and assess the extent to which those sort of rapid-fire policies might be attenuating some of the rates of disease and mortality that we're seeing, particularly in marginalized communities? Anybody happen to know the answer to that, or know of work that might be underway?
Nancy Krieger: What I do know that's underway are surveys right now, trying to find out how people are being impacted by COVID-19. And those are, there's rapid work that's going on in the field for that, whether they are able to also do pre-post intervention, or cross-study designs that can compare what's happening in different places. I think people are trying to get those data, but it's, I mean, it's all just happening so quickly that to try to get that kind of comparison right now, I'm not sure what the outcome would be that you could actually directly compare it to that would be happening quickly enough, given the, if there were complete testing in the population, that would be one thing. But the testing is still so incomplete that that's hard to get at. But there are, the census has just started a new rapid pulse survey that's in the field, going in now, that's getting at the economic disruption to people's lives, getting detailed demographic data, getting at the health disruptions, so I think that's, and that's meant to be representative for all states and then for 50 large, metropolitan statistical areas. And those are gonna be happening and released on a weekly basis, after the first one's released in, I think, about a week or so from now. So I think there are some attempts to get truly representative data that are gonna be able to start allowing one to look at exactly that kind of question.
Mahasin Mujahid: And I'll just add that there are efforts from the National Institute of Health to leverage existing studies, so existing cohort studies, to try to make sure that those can be leveraged to do these kinds of analysis, because they will have data pre and post COVID-19, although we don't necessarily think of a post-19, because I think that sort of the long-term consequences of this are gonna sort of continue on for some time, but at least you will be able to do some of that, when we do some of those analysis of what kind of local resources were in place, areas that had specific, targeted outreach to the communities hardest hit, and other kinds of efforts that are going on, and local efforts will be able to leverage the existing cohort studies to tease some of that out.
Rachel Morello-Frosch: Great. Next question is: the US has had a hard time hearing about structural racism. How can we do better to explain these issues and advocate for change, either as practitioners or scientists? So what's our role in that? How do we do a, yeah, how do we improve that?
Mahasin Mujahid: Well, I mean, I can start. Well, I mean, to be fair, these are really hard conversations to have, and I think, you know, one of the things that has happened is this idea of really emphasizing that this is a system. So, you know, I think a lot of times when people engage in a conversation around racism, they think about whether they, themselves, are an individual actor, right? Are they sort of the perpetuator of this system? And so I think that that's hard, you know, sort of, for individuals to engage with. But the more we help people understand that this system is operating independent of individual actors, that we can sort of really understand how, without acknowledging it, we're going to continue to perpetuate these kinds of inequities in health. And I think the other thing is that we need more representation. I think the more that we have representation in the scientists that are asking the questions and conducting the research, the more we have representation in the people who are deciding on what gets funded and for what research questions, which are legitimate and which aren't, the more we will see this kind of sort of work get implemented in our public health institutions, in our public health funding organizations, and so, you know, sort of the accompanying piece of this is really making sure that we have equity in who is at the table, not just in terms of what we study.
Rachel Morello-Frosch: So in the few seconds we have left, Nancy, do you have any final thoughts?
Nancy Krieger: I just think building on what Mahasin is saying, I mean, yes, it can be hard to have the structural, but that gets back to why: who's benefiting from the ways things are, and what does it mean to have the status quo change? And that's threatening to people who don't want it to change, so we're seeing that play out. And I think that what epidemiology can do, in particular, is that there's two parts of the narrative that are really crucial. One is there remains an ingrained belief that races are biologically distinct. And trying to refute that, you know, it's a generation after generation task, but the science, with each little advance of science, if there's another new thing that can be measured, people will try to say there's a racial difference behind it, but how do you start to show that the social exposures underlie what are then embodied and become differences that you can see, across different social groups? That's a really crucial part of the conversation, because to the extent that people continued to believe that, quote/unquote, races are biologically real. Rather than: the impact of racism is biologically powerful. You have a problem in setting what the basis is for action. Because it's just like: those people are that way. And so I think that that's a really important piece of it, and the other is: showing the evidence in the people's bodies, and individual actions do matter. They're shaped by what the structures are. If there's a permissive structure, as we've seen in the past few years, there will be more hate crimes committed. If there is a sentiment that comes from the top politicians and the White House that it's fine, then you see more of that behavior. So, linking the behaviors that people do actually exhibit, we're always simultaneously individuals and members of our society, they're both true, and the epidemiology can help bring out that context that's shaping what I, ah, the likelihood that individuals experience adverse conditions, and adverse exposures ranging from interpersonal to how they're shaped by what the structures, in which, through people live their lives. We all live our lives, no one's outside of that.
Rachel Morello-Frosch: Thank you; I think that's a great place to end. I wanted to just turn it over to Michael Lu again, who can let us know what our next panel is. But I wanted to thank our panelists for their time. Dr. Nancy Krieger, Dr. Mahasin Mujahid, and Dr. Corinne Riddell. Michael, do you want to let us know what's coming up next week?
Michael Lu: Yeah, so thank you Rachel, and please join me in thanking our panelists for this truly enlightening conversation. Now, this is the kind of conversation that our society has a lot of difficulties having right now, but one that we must have if we're going to truly advance to health equity and social justice. If you found today's conversation worthwhile, I'd like to invite you to join us at our next public event on the impact of COVID-19 on our healthcare delivery system, next Monday at noon. You can find more information on our website at PublicHealth.Berkeley.edu. Thanks for joining us today, and take care and stay safe, everyone.
Rachel Morello-Frosch: Thank you.