Early on in what would be a run of visiting twenty-six states in just July and August, Irum and I frequently departed our hotels at 4 a.m. to the sound of downshifting semi-trailer trucks and the rattle of moths tap-dancing in the shroud of lights illuminating jam-packed parking lots. We hoped that our difficulties in finding available rooms and places where we could dine outdoors more safely and where more people than just the servers and other essential workers were masked would be the exception rather than the rule.
That first trip out of the Washington bubble and into the hot zones was both instructive and daunting. Irum and I armed ourselves as best we could with masks and disinfectant wipes. We had to be careful not just about our health, but also not to let our presumptions rule. Despite wearing masks, we had to be open. We had to be willing to learn. Some of what we learned was alarming. Some proved enormously useful in helping us shape the next approach to mitigation so a state could stay open safely.
We were on high alert. Mobility data showed us that though people in the DC area had decreased their movement by 60 percent, elsewhere—in Texas in particular, but also in Arizona and New Mexico, where we finished up our three-day swing—they were on the move. This meant they were carrying the virus from place to place, many of them unaware of what asymptomatic spread was. A few were heedless, unwilling to listen to anything they were told about the danger they were in or the danger they posed to others. We wanted to understand better: Had they not gotten the message? Had we not made it clear enough? More likely, we presumed, it was a combination of the two.
Too often, public health officials rely on reputation and scaling up the urgency with stronger adjectives or increased volume to change people’s minds and behaviors. After years of working on the ground in communities and among the marginalized, we understood that meeting them literally where they lived and understanding the context of their lives would be essential to our moving forward together. Talking down to people through oversimplification, alienating them by finger-pointing, never works. It may feel good in the moment, when you are frustrated, but it makes the job much more difficult in the end. I had also found that, no matter where I was in the world, no matter the level of formal education, people were smart and had great common sense. They could understand the concepts and the nuances if you took the time to speak to them in clear language. Sometimes, as scientists and doctors, we like to make things complicated to sound really smart and communicate that we should be listened to because we are smart. This attitude and approach completely misses the mark. We should always strive to deliver a clear message, one that can be understood and acted upon by everyone.
We needed to keep everyone on our side because we had ideas we needed them to try. Throughout my months in the White House, I’d continued researching the virus, scouring the internet, reading scientific journals, and speaking with colleagues from around the globe. Irum had been doing the same, and she had discovered an unlikely but critical new model. We agreed that it could potentially serve as another way to actively mitigate and impact community spread without a shutdown.
Dr. David Rubin directs the PolicyLab at Children’s Hospital of Philadelphia (CHOP). He is also a professor of pediatrics at the Perelman School of Medicine at the University of Pennsylvania. Rubin and other members of the PolicyLab staff conduct population and community-partnered research with a focus on the needs of high-risk children. As part of that work, they had produced a highly interactive Covid-19 forecasting model at the most granular local level. They would eventually use it to prepare school reopening guidance and in-person schooling recommendations.
Rubin used precision tracking of transmission and test positivity rates similar to what Irum and the rest of my data team were doing. From that, the UPenn/CHOP team independently produced a forecast that closely matched what we were seeing and projecting. Irum contacted him, and Dr. Rubin created models for us based on various assumptions we provided. We asked him to model the impact of a statewide mask mandate, significant expansion of outdoor dining while reducing indoor dining to 25 percent, and closing standing-room-only bars. We also asked that a model reflect keeping retail spaces open as long as masks were worn in all indoor public spaces. His models demonstrated the impact of these measures on the replication rate of the virus. Simply put, to be under R1 is a condition where every infected person infects less than one other person and community spread contracts. Actions taken to decrease viral replication to a point below R<1 results in controlling community spread.
This combination of mitigation measures, which spared many aspects of the economy, could very well be nearly as effective at preventing community spread as a full-scale stay-at-home-type shutdown. It had the impact of a “shutdown” but wasn’t one. The president had given me that one directive: Do not shut down the country again. I believed we had charted a path forward that could work now (and, critically, in the fall) while meeting his directive.
The UPenn/CHOP model worked in theory. Now we needed to implement it at an actual population level and measure the impact on viral community spread. Many in public health preferred small-scale, controlled experiments, but a real-world implementation was more revealing because many more variables would be tested. We needed to get at least one governor to enact these mitigation measures statewide so we could study their impact in both urban and rural counties. In anticipation of our state visits, we had already laid the groundwork for this on a governors’ call by having David Rubin speak to all the governors about his model and theories.
I believed the UPenn/CHOP model was one the states could use and everyone could follow. If it proved successful with one state, it could be effective as a template, an adaptable way to stay open safely, preserve both economic interests and public health, and ensure that America’s children were in school in person full time in the fall. Every day, I carried with me the deep concerns about children’s mental and physical health and their educational progress.
I was certainly aware that Irum and I, as representatives of the federal government, might not be welcome in the states, that our visits might be perceived as federal interference in their territory. Thankfully, my fears were unfounded. What I’d sensed while in Washington—that the states needed, wanted, more on-the-ground federal support—was reiterated a thousandfold by some governors and health officials. But the states didn’t want just support. They wanted a clear and open dialogue in person to ensure the federal government was hearing them. They also wanted clear guidance like we’d created with the Opening Up America Again framework. They were dismayed that the Trump administration hadn’t shown better leadership in communicating clearly, and they needed the president and others in the administration to model public health mitigation—to wear masks, to tell the truth about the importance of testing, to encourage Americans to stay the course for what was to be a long haul. In the weeks to come—whether it was a red state, a blue state, or a purple state—behind closed doors, when in face-to-face, private meetings with us, governors and other officials repeated some variation of this desire. Their pleas drowned out the sound of the semi-trucks and the fluttering moths. Those voices were what I heard as I fell asleep at night and what I woke up to in the morning.
GOVERNOR GREG ABBOTT OF Texas was our first stop. Flying out of DC on Air Force Two, with its standard airplane cabin preventing physical distancing, I reviewed the data on Texas and my previous interactions with the governor. Throughout March and April, he was among a group of governors—Phil Murphy of New Jersey, Gina Raimondo of Rhode Island, Michelle Lujan Grisham of New Mexico, and Doug Burgum of North Dakota—who frequently, during and outside the weekly gubernatorial conference calls, asked directly for additional guidance and for specific supplies to meet their needs.
Abbott had made many of the right moves in Texas—declaring a state of disaster, limiting dining at bars and restaurants, closing gyms, and restricting social gatherings to ten people. His state had adopted the White House reopening framework guidelines and had worked with me to tweak them as necessary. Governor Abbott adopted the three-phase gated criteria program we’d devised, and it was initiated in Texas on May 1. Like most states, Texas had ridden the waves of the virus, its crests and troughs. In mid-June, the state had eased indoor dining restrictions, and most businesses were open. Just prior to our arrival, the governor had to order bars to shut down and restaurants to once again restrict capacity to 50 percent, the lowest yet. He also banned any outdoor gatherings of more than one hundred people, unless the organizers got local government approval. Texas was experiencing the post–Memorial Day summer surge and was the first state to rescind reopening measures. Abbott was under a lot of heat from both sides—for not doing enough, for doing too much.
Texas was stuck in this feedback loop. We offered Governor Abbott and Texans a way out.
At the start, the governor seemed skeptical that doing three rather simple things—universally masking, restricting indoor dining occupancy, and allowing only smaller social gatherings—could make such a dramatic difference.
“We’ve done these things,” he told me. “We can’t seem to make any real headway. It’s like we’re on a roller coaster.”
“You’ve done two of the three things, Governor,” I told him. “Without mask mandates, you lessen the positive effects of the other two.”
He frowned. “Masks? Nobody seems to agree on that one. It’s a real hard sell. It’s not going to make too many folks happy.”
“It will keep more folks alive,” I said. “That should make everyone happy.”
I went through my explanation of the role silent spread played in the pandemic and said that, short of doing sentinel testing of specific groups, masks were the most efficient means to stabilize the situation. Abbott was tired of the back-and-forth; so were many Texans. Without a clear mandate from the White House and the CDC on the efficacy of masks, he had chosen not to mandate them. Considering what the UPenn/CHOP model demonstrated, and eager to do the right thing, something that would preserve economic interests and public health, he told me he would consider taking some masking measures.
Governor Abbott recognized that his state was in trouble, and on July 2, a few days after we left Texas, he issued a mask mandate. He devised a sensible, creative solution that took into account Texas’s fierce history of independence: the mandate would be statewide but would exclude counties with twenty or fewer Covid-19 cases. This implementation of mask mandates based on the degree of community spread was a data-driven, tailored approach that we learned from him. The result of this approach was that it seemed less arbitrary to the state’s residents. As was true in many places, many of the moves Governor Abbott made were viewed in various quarters as too little, too late, while others saw them as overreaching overreactions. As for me, I was glad to see that he’d added the third element of the UPenn/CHOP model: a modified mask mandate would help us see how effective the three measures could be.
Next, we had to wait for the results. Cases in Texas peaked around July 15, consistent with the masking mandate being added to Abbott’s decreasing indoor occupancy and his clear public messaging. The UPenn/CHOP model was working! We would carry that message forward.
Later, as the governor was being attacked on both sides, either for being late to the game or for overstepping his authority, one state representative tweeted that Abbott thought he was king. But the governor was living out a dichotomy that existed in so many places. In the political and social arena, striking the right balance was possible, but it would take considerable time and a healthy dose of commitment from leaders and the general public to stay in that zone.
As would be true for many states, in the months that followed, state leaders would need to continue to engage in dialogue with those of us at the federal level, so that we might jointly analyze the data and mitigation efforts to review their impact and decide on next best steps. The task force continued this bidirectional learning until the last governor’s report was issued, on January 18, 2021. We provided the governors with clear state-specific recommendations based on their local county data. More important, we listened to each and every governor, independent of party affiliation, and we learned from each and every governor, independent of party affiliation. This is what the vice president believed was critical, and we came to understand why.
Governor Abbott would remove much of the statewide mitigation in the spring of 2021. Yes, he should be held accountable for his decisions. But the federal government should as well. The agencies and the White House didn’t stay directly engaged with every red state or blue state outside of the governers’ calls. They didn’t continue the dialogue we had initiated so that data and issues could be reviewed together. Without that dialogue, neither side was listening and learning. The federal government didn’t revise approaches or provide direct state-by-state recommendations. The virus and the situation on the ground continued to evolve.
Governors require constant engagement and support. They need a clearinghouse of effective strategies that worked in other states. These are lessons we need to carry forward to be ready next time.
We learned two things from the Texas experience: First, Governor Abbott’s decreasing maximal indoor dining occupancy to 25 to 50 percent of the state’s existing fire code occupancy figures allowed for public-sector enforcement and consistent application. And second, closing the “bars” in Texas wouldn’t have done much good. Nearly three thousand establishments that in other states might be classified as “bars” are, in Texas, classified as restaurants, better known as “roadhouses.” As a result, if we used the term bars when referencing closures, the owners of these roadhouses could make the case to health officials that We’re not a bar, so that doesn’t apply to us. It may seem obvious that “bars” refers to places that serve alcohol, but this linguistic and legal distinction was one example of how our general guidelines were being interpreted and applied on the ground. During our weekly governors’ calls, there wasn’t time for any governor to raise unique concerns like this. For this reason, such critical nuances escaped us. In person, we got the message. The bar-versus-restaurant distinction alone made our Texas trip worth it. It proved that being on-site could yield positive results.
In Texas and elsewhere, we realized that we couldn’t always trust what we were being told. Some state officials were reluctant to report bad news to the governor or to Washington. Also, some were using bad data as a source of such reports. The use of state averages rather than local county numbers was an ongoing problem. State averages provided a misleading picture of where the hot spots lay. While, for example, Indiana might have had a low case incidence rate at one point, test positivity in Brown County or in the zip code 46135, in Putnam, was exponentially rising. Working at this level of precision was a major key to an effective mitigation response.
The tyranny of averages played a role in a less-than-proactive response. While reports at the federal level indicated that our supply management was in great shape, we found out that some states were lacking much-needed tests and testing supplies (swabs, tubes, extraction media, pipettes), remdesivir, ventilators, and other crucial items. This information was helpful but disheartening.
Personally, we were finding solutions to some road trip issues. We learned where there were safe, secure places to pull off the road to change clothes discreetly. We even figured out how to do so in or near the car—wedged between the open front and passenger doors and, at least a few times, behind trees and Dumpsters. In minutes, we would go from car-comfortable stretchy pants to full business dress, stockings, and heels. And we developed a new appreciation for fast-food drive-thru windows. We also found that those working drive-thrus were a good source of information about how a particular community was doing vis-à-vis mask use and other aspects of mitigation.
We continued to read about the negativity and hostility flaring up around the country over masks and shutdowns, but we saw no real sign of it, and none was directed toward us. We were moving around incognito—after all, we were masked—and that helped. Not being recognized felt right. I’d later be recognized, even while masked, when I ran into a CVS for mascara. That experience unsettled me. I wasn’t used to that kind of attention, and though the clerk was very friendly and helpful, it still felt odd to be recognized. I’d be spotted again a few times over the course of the next few months. Once, at a filling station/small general store in Oklahoma, I spoke with two men from Florida who were on their way to Idaho to fly-fish. We talked about being really careful, as the virus was moving north. They thanked me and our messaging for having convinced them to drive rather than fly.
After our meetings were concluded in Texas, we set out for New Mexico. I wanted to go there specifically because it wasn’t in crisis, yet two of the states it bordered, Texas and Arizona, were. It was important to see what New Mexico was doing differently. The state had far fewer per capita infections than the other two, and though I’m normally not keen on averages, that did say something big-picture about how it was responding to the crisis. From the governors’ calls, we already knew that Governor Lujan Grisham had been very assertive. She had a strong belief in the public health value of testing and mitigation. (She also enjoyed being a Democrat in a state with a Democratic legislature.) In fact, Irum and I wouldn’t be allowed to meet with the governor personally unless we tested negative for the virus. In spite of how lax the precautions had been in the hotel during our overnight stay in Amarillo, our subsequent tests both came back negative, which was reassuring but too early to mean anything.
Governor Lujan Grisham, her staff, and the state’s health officials greeted us warmly. In our discussions, we got into some of the nuts and bolts of how the governor had advocated for and put in place restrictions on indoor dining and had tried to keep outdoor dining options available. Where other states had not, she had been able to keep retail businesses open with aggressive distancing and masking. She had provided options to the citizens of her state, and they had responded well. I made note of New Mexico’s success, particularly with its tribal nations. Core to Lujan Grisham’s approach was the constant expansion of testing. She had been the first governor to call me to ask for more tests back in March, and those requests continued as she used the tests to find the cases early to stop community spread. New Mexico’s population is diverse, and its health officials had paid close attention to the state’s unique needs and had supported all groups. Though she wasn’t aware of the UPenn/CHOP model mitigations by name, Lujan Grisham had employed them.
Happy to have an example of how they worked outside a computer-generated model, we got on the road for Phoenix.
While the people we encountered over those first two days were nearly without exception friendly and courteous—even the Texas policeman who issued Irum a warning and not a ticket for her spirited driving—this didn’t mean some people weren’t angry. After completing early meetings in Phoenix, Arizona, I had a scheduled call with California governor Gavin Newsom at 5:30 p.m. I stepped out of the car into a blast furnace of heat. The governor wasn’t able to get on the line, but his chief of staff and senior health advisor harangued me for the first ten minutes of an hour-long call, vehemently complaining about the lack of leadership from the White House. In that call and through future state and local meetings, I would hear multiple examples of Trump’s faults and the damage they had done. My talk with Governor Newsom’s people was merely a preview:
“If only the president would just talk about masks.”
“If only the president would talk about the importance of widespread testing.”
“If only the president would just—”
I’m sure my responses did little to assuage them.
“I understand.”
“I know.”
“I, and others, are trying.”
I sensed that they needed to vent their (or the governor’s) frustrations. Once California was in a better place, they would be more receptive to the message I was formulating.
As painful as it was to feel the heat of the day and other people’s displeasure, it helped solidify my vision: We had to travel to literal and figurative hot spots, where the weather and the virus were heating up. It was summer, and Arizona and the Southeast, where we would go once this three-state swing was over, were in the Goldilocks zone: too hot. With the extreme heat, people were being driven indoors, gathering together into air-conditioned spaces—optimal conditions for viral transmission.
The day after the call with Governor Newsom’s staff, Irum and I walked into an office building in Phoenix to present our three-pronged mediation approach to safely reopening and staying open to Arizona governor Doug Ducey. He was open to a discussion of the UPenn/CHOP model.
“That sounds good in theory. But I can’t see how it can apply here. Arizona’s unique. We’ve got an aging population with lots of retirees. We have tribal nations at great risk. We’ve got a couple of major metropolitan areas. We’ve got people living in small pockets out in the desert and in the mountains. ‘One size fits all’ isn’t going to work here. We’re too diverse in our geography and in our politics.”
He wasn’t alone in this perception. Fortunately, Governor Ducey and his health team were willing to listen to our proposal. They reviewed the charts and graphs. Dr. Cara Christ, director of the state’s Department of Health Services, was bright and hardworking, and she understood what we were trying to do. We learned a lot from Arizona’s team, and Dr. Christ would become a critical partner in the months to follow.
We also met with Arizona’s county health officers and community groups along with the governor. They had already started outreach in key Black and brown communities. The county health officers had a clear-eyed understanding of their communities and had developed a meaningful partnership with them.
One consistent claim—in Arizona, Texas, and California, and from many mayors in those states and elsewhere—surprised me. Many told us that outsiders had imported the virus to their locale. This wasn’t a case of xenophobia, or of the despicable anti-Asian sentiment we had seen while watching media coverage. Instead, it was a variation on the “over there” problem. Leaders felt they were being responsible in their county or state, but that the state next door, the county next door, wasn’t being careful, and now their area was failing due to others’ carelessness. People crossing the border are the ones responsible for the infestation of the virus, not us. It wasn’t here. Someone brought it here. In Southern California, someone said to me the problem was people from Arizona coming to the state to go to the beach. Texans told me it was the people from Louisiana, or Texans who had traveled to Louisiana (especially, to New Orleans), who had brought the virus back to Texas with them. Somehow having the virus in their state seemed more palatable if it was someone else’s fault. My community is good. Your community is bad. And now you’ve made our county sick. It was brought here.
In Arizona, the governor and his public health officials were willing to implement the full mitigation measures that the UPenn/CHOP model showed would have an impact in two weeks. We made this point with them: We weren’t imposing a strict plan, but rather an easily modifiable one that the governors and public health officials could revise and make their own. Having heard from Vice President Pence about how the governors thought, operated, and succeeded, this “You craft this, you own this, you have responsibility for this, and we will support you” approach proved effective in selling this message to the states. Conversely, we also gave the governors an out. If the measures we urged them to take failed, they could always blame the federal government for the ideas we’d brought to their state.
Ducey’s actions were data-driven, as were those of many other state leaders. Governors who had come to public office after success in the business world immediately understood the graphics and the need to use data to make decisions. They had been doing that their entire business careers. I saw this data-driven approach from Governor Jim Justice in West Virginia, Governor Tom Wolf of Pennsylvania, and Governors Burgum and Ducey.
Governor Ducey reviewed the model and the data with his health advisors. Given the demographics he’d cited, the mitigations we were championing were a good fit for his state, especially with the large numbers of vulnerable Americans living there—older Americans and those living communally, Native Americans on tribal lands, and those living in long-term-care facilities. Ducey’s progressive, analytic approach allowed Arizona to thread the needle between public health and economic interests, creating a road map we could offer other governors. Using that map, they could select the route they wanted to take to control the viral spread in their state.
In Governor Ducey’s “We’re different from other states” remark, I saw another objection I had to counter. It is always hard getting people to understand that two somewhat contradictory things can be true simultaneously. Yes, a state could be unique, but no, it wasn’t the only state going through this. On this first trip and later, when in a given state, I found it helpful to point out that what they were experiencing was what other states were also going through or had already gone through. We’d then show that state’s leader what other governors were doing and how it was working. This message of “You’re not alone” helped offset the sense of failure and potential resignation governors might have felt in having hot spots in their state. This was a bit tricky at the beginning. Because the Northeast corridor had been the site of the first and most severe outbreak, the March–May 2020 surge, the virus was perceived as a big-city, “over there” problem. The places we were visiting weren’t packed metropolitan areas, and the thinking was It can’t happen here.
This wasn’t true. Covid-19 was an urban, suburban, and rural problem—and very real. But we had to prove this.
Because they hadn’t seen their own March/April surge, too many communities believed they were immune. This was a consistent theme in our conversations. All we could say in response was that, regardless of how the virus had gotten there, it was now time to concentrate on limiting and eventually ending its spread. In the end, this brief, busy, and nearly sleepless first trip accomplished what we had hoped it would. We remained concerned about those who were truly on the front lines, not those in power who had chosen to place themselves on the sidelines. If too many people put their heads in the sands of denial or anger or bargaining, instead of focusing on what we could do next, we’d all be grieving over what might have been. That was a place I never wanted to travel to.
On the flight back to Washington, I wrote to Jared Kushner, pulling no punches as I spelled out what the word was out on the streets about the White House leadership: The governors appreciated the supply support, but all of them wanted consistent federal messaging on masking, testing, and gatherings. This wasn’t coming only from me. It wasn’t coming only from Democrats or those on the left. It was coming from Republicans and those on the right. Left, right, and center—the persistent chorus contained voices that needed to be listened to.
In community meetings during that trip, I had also seen—along with frustration over messaging and, at times, anger—real concern. This was especially true for the women who gathered to listen to us and share. When I mentioned how important masks were, the women present nodded their heads. I saw them glancing at one another and then at their husbands. The first look was one of satisfaction at hearing that their concerns were legitimate. The second was a “See? Why won’t you listen?” look. Across the board, moms, grandmothers, daughters, and sisters were concerned and anxious. Like me, they had vulnerable family members—sometimes a spouse, sometimes a child, someone with Down syndrome or severe asthma, sometimes an older parent. Although many were silent, their eyes held this worry.
I knew the feeling. It’s hard to know that you’re right but that it doesn’t matter.
The entire executive branch needed to understand what was needed. We needed to listen, and to respond with support. Guidance and words on websites weren’t enough. To truly support implementation, we had to translate those words into actions in deep partnership with state and local leaders. We couldn’t simply issue guidance to the states and then take a hands-off approach. This trip showed me that we could provide the states with clear examples of what was working on the ground with specific populations. We needed to gather more of this kind of “data,” and we needed to share those insights and successes during our weekly governors’ call and the written governor’s reports, and that’s exactly what we did.
Irum and I were energized by the visits. At the state and local level, in counties and on tribal lands, solutions existed. You could see and learn from them not from DC or Atlanta, but only when you were out in the communities where good people struggled to follow complex federal guidance. You couldn’t just issue a statement and expect it to be understood and followed. In order to adapt your communications approach to the people to whom you’d targeted your message, you had to meet those people where they were, figuratively and literally. That was the only way to know if you’d made a bull’s-eye or were off the board entirely. This strategy, which had served me well around the globe, was important here in America, too.