Tim Hoff: Welcome to another special edition of Ethics Talk, the American Medical Association Journal of Ethics podcast on ethics in health and health care. I'm your host, Tim Hoff. This episode is an audio version of a video interview conducted by the Journal's editor in chief, Dr Audiey Kao, with Dr Mollyann Brodie. Dr Brodie is the Executive Vice President and Chief Operating Officer at the Henry J. Kaiser Family Foundation. She joined us to talk about the science of polling, public opinions on COVID-19, and politically divergent views on policy options to achieve universal health care coverage. To watch the full video interview, head to our site, JournalOfEthics.org, or visit our YouTube channel.
Dr Audiey Kao: Good afternoon, Molly. Thanks for joining me on Ethics Talk today. [music fades out]
Dr Mollyann Brodie: It's so great to see you. It's been a while.
Brodie: Nice to talk to you. I'm excited to be here.
Kao: I hope things are well with you.
Brodie: As it can be expected in the world [inaudible; cross-talk].
Kao: That's true. That's true.
Kao: So, Molly, as you know, this year marks the latest decennial effort to count every person living in the U.S. and the five territories. All of us are expected to do our civic duty and complete the U.S. Census questionnaire. As opposed to a census, however, polling aims to understand public opinions by collecting data from a small sample of the overall population. I know one of your other responsibilities at the Kaiser Family Foundation is overseeing the Public Opinion and Survey Research Program. So, before we delve into public attitudes and knowledge about COVID-19-related matters, can you provide our audience with a quick ABC of public polling?
Brodie: Yeah, I'd be happy to. First of all, let me do a plug and say, please all, fill out your census forms, because that data couldn't be more important to so many things that happen in our nation. I want to use a quote from Winston Churchill to start. He said once that, “No one pretends that democracy is perfect or all-wise. Indeed, it has been said that democracy is the worst form of government except for all of those other forms that've been tried before it.” And I think that perfectly sums up exactly how people feel about public opinion polls. It's a thing that everybody loves to hate, but we can't live without it. And at the end of the day, it is still the best and the most reliable scientific way to find out what a population you're interested in thinks, experiences, or believes. And it's really based on the beauty of statistical sampling and methodology. And because of mathematical statistical formulas, you can talk to, as you said, a small number—it's usually about 1,000, 1,200, 1,500 people—of any given population and be able to speak with a lot of confidence about the views and experiences of the whole population.
Brodie: So, instead of talking to 300 million adults, I can talk to 1,200 adults and do a pretty good job, within plus or minus three percentage points, of telling what people think. And it's not different than, I'm imagining much of your audience is physicians, and you do blood samples all the time. You don't have to drain your patients of all of their blood to find out their levels. I like to, when I teach about sampling, I use my soup example where you're heating a pot of soup on the stove. You don't have to drink the whole pot of soup to know if you're going to serve it. But what you do need to do is do a really good job of that sample you pick, right? So, if I take a bite of soup from the very, very top or from the very, very bottom of the pot, or let's say it's a vegetable soup, since I'm a vegetarian, and I only taste the carrot, I don't taste the potato, then I might make a mistake when I serve my soup. It might be burned. It might be too cold in some places. The vegetables might not be hot. But if I do a good job on my methods, if I do a thoughtful job about storing my sample and randomly choosing it, I can be pretty certain that if I take a taste, it's good to serve.
And it's exactly the same thing with public opinion polls. If you do your methods right, if you really well define your population, and do a good job selecting your sample, then we have a lot of confidence that— And there's a few other things. Like we need to be a little thoughtful about the questions we ask and the order we ask them and a lot of other science of surveys. But if we do all that right, then we can represent the populations you care about. And you just have to define that right. Is it all adults? Is it voters? Is it those with chronic illness?
Brodie: Is it the uninsured? Right now is it people who have COVID? Is it people who have tested positive for COVID? And so, you just want to define your population, your methods right, and then you can really, really say a lot about what that whole population would say if you could talk to them all.
Kao: Yeah. No, I appreciate your short primer. And I'm not sure I'm going to look at soup the same again after your analogies.
Brodie: There's always other people use bourbon tasting or wine tasting.
Kao: Uh-huh, uh-huh. Right.
Brodie: There's all sort of ways you can go about it, but yeah.
Kao: Sure, sure. So, to extend the points you've made, what should health policymakers know—or even the public at large, for that matter—know about what distinguishes good scientific polling from poor polling? And what polling strategies and technologies should be used to help solve big problems like we're seeing in this current pandemic, such as racial and ethnic health disparities?
Brodie: Yeah. I think it is definitely harder and harder for sort of a casual observer of polls to be able to distinguish a really high-quality, very representative one from maybe not so representative. There's lots of reasons for that. The barrier to entry to do polling has changed dramatically. There's lots of new technologies. And so, I think that would what I would say to everyone listening in now is you have to be a good consumer. And a good consumer looks for transparency. So, the more information that a poll provider, a pollster, is willing to tell you about how they selected their sample—how they defined their population, was it a probability-based survey or non-probability-based panel? How did they weight the data? What were the questions? What order were the questions in?—the more confidence you can have about what kind of conclusions you can draw from that data.
And I would say there's fit for purpose, right? Not every set of data that we need to collect has to be the multibazillion-dollar study, right? Because for some purposes, having a decent sample, a large sample so can spread out your errors with some reasonable questions, you can learn a lot. But for certain public policies and for certain conversations, and particularly depending on who you're claiming it's representative of, then these enhanced efforts to make sure that you're doing the most random probability-based project you can do matters more.
Brodie: And so, a couple other things I would, tidbits to make you a better consumer: I would never draw a strong conclusion based on one question, right? It often takes a handful of questions that are all going in the same direction or indicating the same sort of conclusion that you'd want to be able to draw a strong conclusion from. I might look at multiple polls and see if there's consistency around what you're learning. Sometimes the polls that get the most attention or the poll finding that gets the most attention, it makes news because it's an outlier. Well, you want to question, well, why is that an outlier? [chuckles]
Brodie: If it's different than everybody else's, is there a reason it's different than everybody else's?
Brodie: And so, those are some of the things. The other thing that I would just say is that it does matter how strong the interpretation is. So, depending on how the person who's presenting the poll, you know, most of the listeners probably know a lot about statistical significance, right? Statistical significance and substantive significance are not necessarily the same thing, particularly when you're talking about opinions and beliefs.
Brodie: So, even if it's statistically different, it's a four-point difference between men and women on whether they are staying at home as a result of COVID, if it's 83 and 87 percent, well, you know what? Those are really large shares of both populations.
Brodie: And I might think that it's not so substantively meaningful that that's a statistically significant difference.
Brodie: So, there's a lot of ways you can be a better consumer of polls. One plug that APOR, American Association of Public Opinion Research, has a transparency initiative. And organizations that are members of that transparency initiative have agreed to be more transparent about their methods. Now, sometimes that's a quick clue, but at least it's an organization that has been willing to be very transparent about how they do their work. And that goes a long way in my book to at least being able to make good judgments about things.
Kao: Yeah, I appreciate that. I think your points about being a skeptical consumer of polling is also quite important. If I can just follow up on some of the points you made about interpretation.
Kao: As you just described various aspects that distinguishes good or accurate polling data from inaccurate polling data, but the relevancy and trustworthiness of poll data may be more a function of how the data is interpreted and presented. Is this distinction correct? And if so, why is this important in health policymaking?
Brodie: Yeah, I think it is. You know, as I said, I think it's a gray area, and it's hard to grade a poll as really good, really bad. You can sort of [audio drops] the sides out. But that middle, there's a lot of people who are doing pretty decent work, given their best methods. And it's a little bit gray in there. But I do think that interpretation and how, sometimes we use the term “spun,” “spin,” how it's being…. How careful are people being about their interpretation of what they found does go a long way, I think, to help give you confidence about what conclusions you can draw.
Brodie: And also the purpose of the poll. I mean, there's different purposes. I mean, the KFF, we do polls for a number of reasons. Mostly we do them because we want to give voice to people in a political system who might not always have a voice. The uninsured, people with chronic conditions, people of lower income, of disadvantaged backgrounds: we want to make sure that their voices are represented on the political [inaudible; cross-talk] tables.
Brodie: We also want to make sure that people's real-life experiences are part of a policy debate, and it's not just always opinion or political outcomes that are being counted and paying attention to. But what I would say about any interpretation of polls is that as much as respect as I think you have to have of the public and of their true beliefs and their opinions, it doesn't necessarily mean the public's right.
Brodie: People are still— It tells you where leadership is necessary. That's how we interpret our polls. It tells us where public education is necessary.
Brodie: It tells policymakers and journalists where their audience is and where they have to start with their audience and be respectful of the audience's beliefs and opinions on the matter, and then try to help explain why the policymaker or the journalist or whoever believes differently. Why, in a case of a public pandemic, we might not be able to follow what the public says they want, because maybe the public health experts and other folks have facts and information that the public needs to be taking into account. So, I guess in my mind, the most important thing as public health professionals, as clinicians, as experts, is to remember that the data tells you a lot about the audience and about where people are. It doesn't necessarily tell you what the right thing is to do, but it tells you what your challenge is going to be. In terms of being a leader, in terms of being an expert, in terms of being an educator, you'll know what you're facing. And you might have to rethink your message and rethink how you're talking about an issue in order to meet the public where they are.
Kao: Yeah, that's an interesting point. So, I think in this case, the consumer may not always be right.
Kao: And it points out to decision makers at all levels at least what the public is perceiving and thinking about. And that will hopefully shape better solutions, policy or otherwise, going forward.
If we can switch gears and focus more on COVID-related public attitudes. I know that KFF, as you mentioned a moment ago, the Kaiser Family Foundation, and other organizations have been polling regularly about COVID-19. Can you share with our audience some of the more important findings that have emerged from this public polling?
Brodie: Yeah. Wow. I would say they fall into probably about four buckets. The first bucket is something that you just really have to take a big step back and pause and think about the ramifications of this. We have, for the first time, probably at least in my lifetime, had an incredibly shared collective experience. The polling data says that 84 percent of us have had our lives really disrupted. 80 percent of us have done a good job sheltering in place. Three quarters of us have bought or purchased or made masks for ourselves. 80 percent of us recognize that these strict public health measures are important, and more important right now, than necessarily the economic harm that's coming from it. I can't remember a time in 25 years of polling where we've had such a collective experience across our nation.
Brodie: So, I think it's really important to take a moment and remember that. And I think we really did learn that from all of the public polling.
Now, next basket of findings is the persistent and continually growing partisan divide in terms of what should be happening next and the policies and moving forward.
Brodie: And this is very reflective of what we know about partisan polarization in the last decades or so. But we're really starting to see it emerging as much as Republicans and Democrats and Independents agree about the experience so far, thinking about what should happen next, we're starting to see big divergences between Democrats and Republicans in particular. With Republicans, more likely, the thinking is the worst is already behind us. It's time to move on and start taking care of our economy. And that it's really time to open up. It's not universal. It's not the anecdote and the very, the things we're seeing on TV of the protesters. That's still a small share of the public, but there definitely is a partisan divide there.
The third thing that I think the polling has really shown is the extent of the economic hardships that have happened to so many of our fellow Americans.
Brodie: So, four in 10 live in a household where either they or their spouse has lost a job, been furloughed, had their salary cut, their income cut in some way. And while we can see some of that through the unemployment filings, you can't see the extent to what it means for families. So, we've been able to measure the four in 10 percent of that group, four in 10 who haven't been able to put food on their table, haven't been able to pay their bills, have already eaten through their savings. And I think the economic hardship is a really important component of these tracking efforts over time that we and other people are doing to try to see what's going to really happen to these families as this pandemic continues and as the economy hazard does not rebound quickly.
I think the other thing about this that's important, and what we learned through the polling, is that this group, this economically hard-hit group is not any more likely to want to get back to work than anyone else. They really recognize the importance of the shelter-in-place and of the things that we're doing, and I think that's something else that we could only learn through the polling, I think. I imagine going forward, and as the pain and suffering of having no income increases, they may be among the first groups who want to make sure we're going back to work. But the essential workers and the health care workers and this group of people who are really badly economically hurt are as likely as anybody else to think that the shelter-in-place mechanisms are important.
And I would just say the last thing is that we've really learned that Americans understand to a large degree the importance of these public health measures. I mean, back to, in a sense, circling back to my first point, the fact that so many Americans did what they were asked by their leaders and stayed home and follow the rules and protected their families and their loved ones to the best of their ability was really amazing.
Governor Doug Burgum: [from video posted on YouTube] I would really love to see in North Dakota that we could just skip this thing that other parts of the nation are going through where they're creating a divide, either it's ideological or political or something, around the mask versus no mask. This is a, I would say, senseless dividing line. And I would ask people to try to dial up your empathy and your understanding. If someone is wearing a mask, they're not doing it to represent what political party they're in or what candidates they support. They might be doing it because they've got a five-year-old child [voice shaking] who's been going through cancer treatments. They might have…vulnerable adults in their life who currently have COVID, and they're fighting. And so, again, I would just love to see our state as part of being North Dakota Smart also be North Dakota Kind, North Dakota Empathetic, North Dakota Understanding to do this thing.
Kao: No, I appreciate your perspectives. I think that you're right that outside of the Great Depression and World War II with the quote-unquote “greatest generation,” it's hard to think of another time when we've had this shared collective experience you've been describing.
Kao: If I can just follow up a little bit. You mentioned that, in many ways, the glass is eight tenths full rather than two tenths empty.
Kao: But you're also seeing an evolving, growing partisan kind of divergence.
Kao: Do you want to speculate—obviously, given that this year is an electoral year—do you want to speculate about how those public viewpoints will likely play out for the rest of the year?
Brodie: Yeah. That is one of the hardest questions. I think, you know, we're six months out from the election, and usually, we have a little bit better idea of what it's shaping up to be. But there's just nothing usual about this moment in time.
Brodie: We don't know how the campaigning is going to happen. We don't know how voting is going to happen. We don't know what the status of the pandemic will be in November and whether it'll be even safe for people to go out and vote. So, with that said, I do think that we know a couple things. We know that during the primaries and leading up to this stage that the top issues on people's minds were the economy and health care.
Brodie: And we know that that's still going to be the case. Those are going to be the things that we're going to be talking about headed into this election. It is very likely that we're going to be talking way more about the economy than we probably even would've before given what has been happening. So, those will be the topics. I think we've known that President Trump has a very strong connection to his base and that there hasn't been a lot of movement in that base. They support him. They support him in his actions on what he's done throughout this pandemic. So, I think we know that that's still true. I think we know that Democrats have always cared more about health care and are more interested in the issue of health care and that this pandemic has put many of those issues really front and center. So, that may very well help to be a motivating factor for the Democratic base to be out and voting loud. But at the end of the day, the election is determined by who shows up to vote that day. And we just don't know what the pandemic's impact is going to be on voter turnout and on voter motivation across the spectrum.
Kao: So, as we near the end of our conversation today, you had mentioned a few moments ago that pre-pandemic and most likely right now, the issues that will be on the minds of voters when they vote in November, whether by mail or in person, will include health care as a major issue. And so, as you know, the Supreme Court later this year will hear a lawsuit from various GOP-led states that believe that the Affordable Care Act is now invalid because Congress has eliminated the tax penalty for not having health insurance. But this pandemic has exposed many things, but one of those are the consequences when quality health insurance is not afforded to everyone. What does the current polling data say about the public's attitude about different policy solutions, including the Affordable Care Act, to achieve universal coverage? And how should we interpret public opinions depending on what you call or label a policy solution? So, for example, calling it Medicare for All versus a single-payer health care system?
Brodie: Well! So, [laughs] that question basically could take an entire semester to unpack.
Brodie: I have been to hour-long lectures [unclear; cross-talk].
Kao: We'll have to invite you back for part two then.
Brodie: Yeah, it is basically part two. But let me just try to sum it up in a couple points. I'll do my best to keep it short. Wow. Okay. So, the American public has always, it's always easier to agree on goals than it is mechanisms to achieve those goals. So, and that certainly translates into language or how you describe policy too. So, for example, the idea of universal coverage or making sure everyone has access to health insurance is a very bipartisan, universal belief. People agree with that. It's very positive. One of the reasons why more candidates talk about universal coverage, making sure people have access to affordable health care, they're going to not offend anybody. That's all fine. Policy solutions themselves have much more of a partisan tinge to them.
So, the original reason Medicare for All was called Medicare for All, even though I'm sure, as most of you all know, the actual Medicare for All proposals don't even really start with the Medicare program, but it's because the Medicare program was always very popular in, again, a bipartisan basis. So, it was this popular term you could use to talk about a single-payer system, because if you use the term “single payer” or “a national health plan,” you have much fewer people thinking that was a good idea. It was much more associated with sort of a Democratist, socialist sort of message. And socialism is still very, surprisingly, well, not surprisingly, it's an unpopular term. So, the terms do matter. It's a reason, the terms matter so much is the reason Medicare for All became the term to describe kind of a single-payer system. And now that it's become so politicized in the course of the Democratic primary campaign, it's also a reason why Medicare for All is no longer in any way, shape, or form favored by Republicans at all, right?
Brodie: Because now they see it as a Democratic, very liberal Bernie Sanders sort of proposal. But anything that involves choice or option is considerably more popular. So, as soon as you do anything about a public option or a Medicare buy-in or Medicare for certain ages, you get in the 60s to 75-percent range of favorability. So, you get a lot more people interested as soon as you start talking about something as an option. So, that's the first point I want to make.
The second point is you asked about the ACA, and the ACA is also a very complicated policy proposal because it isn't one thing, right? ACA was a set of a variety of policy changes. And when we ask about each of those policy changes in and of themselves, they're almost always universally appealing and liked by almost equal shares of Democrats, Republicans, and Independents. But the ACA overall has always been seen in a very political vein. It's always been supported by Democrats way up here at the top of my table. I've asked about it every month since 2010; so, about 90-100 surveys. We always get about 70 percent of Democrats favoring it or more. And you get only about this share of Republicans favoring it, about 20 percent. It's completely bimodal in terms of the distribution of who favors and opposes the ACA.
Brodie: And that's because it was seen purely as whether you liked or didn't like the Obama administration.
Brodie: So, while Republicans are as for pre-existing conditions and having kids on their parents' plans, and no cost strings for preventative medicine, those are all things that are very popular. Even providing subsidies for people who can't afford it through the marketplaces and having Medicaid expansion, those are also popular among, for Republicans. But the ACA as a whole is a very unpopular concept for Republicans. When it comes to the Supreme Court case, you still see that Republicans want the ACA repealed, and they are happy for the Supreme Court to make it illegal. But they don't want pre-existing condition protections to go away. They don't want many of the popular provisions in the law to go away. And that's exactly the problem that the Republicans got into in 2017, 2018 when they did the repeal replace debate. Because they have a lot of support among their base for repeal and replace, but not for the plans that they were proposing because they were all very scary to people because it's taking away some very popular provisions.
What also happened because of the repeal and replace debate is that it revved up Democrats' support for the law. And it was sort of a threat that for the first time, they felt really threatened that the ACA really could go away. But we actually, for the first time in all of our measurements we saw, we used to see when you look at the data overall, sort of half favored, half opposed. But ever since the Republican repeal and replace debate, we've seen more favor than oppose by a chunk, and that chunk is more Democrats and more Independents sort of rallying around the ACA as a result of the threat of repeal. And that has been maintained ever since that debate.
We also know that in 2018 elections, the threat of losing pre-existing conditions was an incredibly powerful campaigning piece for the Democrats that helped them take back the House with so many seats. And so, I anticipate that as we go into the election, depending on the status of COVID, there will be candidates who will use the threats against the pre-existing condition protections and the other protections as part of their talking points when they're on the stump.
Kao: Yeah. Well, I appreciate your final thoughts.
Brodie: [unclear; cross-talk].
Kao: We're going to have to bring you back maybe after the November elections to do kind of a postmortem, so to speak.
Kao: But I want to thank my colleague, Mollyann Brodie, for sharing her expertise and insights with our audience today. Molly, thank you for joining us today on Ethics Talk.
Brodie: Thank you. Thank you for having me.
Kao: For more COVID ethics resources, please visit the AMA Journal of Ethics at JournalOfEthics.org. And to our viewing audience out there, I want to encourage everyone to complete their U.S. Census questionnaire. Every person matters, and we all count. We'll see you next time on Ethics Talk. [theme music plays]
Credit Designation Statement: The American Medical Association designates this enduring material activity for a maximum of 0.50 AMA PRA Category 1 Credit™. Physicians should claim only the credit commensurate with the extent of their participation in the activity.
Disclosure Statement: Unless noted, all individuals in control of content reported no relevant financial relationships.