COVID-19 and the Culture of American Federalism =============================================== * Emily Pears * Emily Sydnor ## Abstract COVID-19 highlighted America’s federalist structure as the dissemination of pandemic information was frequently left to states and localities. For some citizens, this was a welcome relief from national-level policymaking and political narratives, though others argued that the federal government was failing to live up to its obligations. We identify three reasons for variation in Americans’ trust in information from different levels of government: partisanship, ideology, and state identity. Using data from a representative online survey of more than one thousand people, we demonstrate that each individual characteristic shaped respondents’ trust in leaders to provide pandemic information. Partisanship and ideology played major roles in information trust at both the national and state level, but individuals’ psychological attachment to their state and to the nation also shaped their trust in the federated information environment. * place-based identity * federalism * COVID-19 * partisanship * political information COVID-19 created an information crisis for Americans unlike any other in the twenty-first century. Across the country, citizens became desperate for information about the pandemic, including scientific and medical information about how the virus spread, technological information about the prospect of vaccines, and policy information to better understand local, state, and national responses. Most citizens have a set group of sources they rely on for political information and health-related advice. But in the context of the pandemic, many required new and different kinds of information, urgently. Citizens who generally relied on their family doctor or the advice of local school nurses for health-care-related information were left scrambling to sort through the new-to-them world of infectious disease experts, epidemiologists, and national health agencies. Savvy followers of American politics are used to consuming specific policy information through the lens of the news media, but the pandemic put press conferences front and center and led many more citizens to consume information directly from the president, their governor, or local officials. Many of us not only learned that such things as county public health departments exist, but also learned the names of the directors of those agencies. American citizens, particularly early on in the COVID-19 pandemic, were desperate for information and overwhelmed by choices of whom to trust for that information. That highly salient information environment was also far more federated and decentralized than we have come to expect in modern American politics. Americans could choose to get pandemic-related information from President Donald Trump, the Centers for Disease Control and Prevention (CDC), Dr. Anthony Fauci, their state government officials, their local government officials or national, state or local media sources. Whom citizens chose to trust for pandemic information was, we argue, shaped by their partisanship, their ideological views of federalism, and their own place-based identities. Whom you trust to provide pandemic-related information matters a great deal. As political scientists Courtney Page-Tan, Summer Marion, and Daniel P. Aldrich (2022) show elsewhere in this issue, citizens’ adoption of behaviors designed to curb the spread of COVID-19 was fundamentally shaped by the civic networks and information sources they chose to rely on. The political scientist Elizabeth Suhay and her colleagues (2022, this issue) also show that citizens’ trust in government was associated with either higher or lower likelihood of engaging in protective behaviors based on the information being provided by leaders of those governments. If citizens with greater trust in their state governments are more likely to support increased state responsibility for policymaking, as the political scientists Marc Hetherington and John Nugent (2001) argue, trust in state governments, and the information they provide, should matter a great deal for citizens’ ultimate assessments of the pandemic response and, in some cases, their likelihood of engaging in potentially life-saving behaviors. We argue that citizens’ existing attitudes fundamentally shaped their assessment of whom to trust in the pandemic crisis environment. As research elsewhere in this issue demonstrates, who citizens chose to trust for information, in turn, influenced their health-related decision-making (Page-Tan, Marion, and Aldrich 2022, this issue; Suhey et al. 2022, this issue). We identify three individual-level variables that, based on existing literature, we might expect to influence citizens’ assessments of a federated pandemic information environment: partisanship, ideology, and state identity. We examine the effect of each characteristic on Americans’ trust in a range of national and state-level politicians and agencies to provide pandemic-related information, drawing on survey data from a representative sample of U.S. adults. We find some measure of support for each theory; partisanship, ideology, and state identity all affect Americans’ trust in federal and state officials to provide pandemic information in an environment where federalism is highly salient. To many scholars of political behavior, these findings will not come as a surprise; researchers have found for decades that partisanship plays a significant role in individuals’ political beliefs, including their media habits and trust in government institutions and information (see, for example, Campbell et al. 1960; Green, Palmquist, and Schickler 2002). Existing political science frameworks would predict that Americans’ pandemic information-seeking behavior is driven by their partisan identity and by national political narratives. America’s pandemic response relied heavily on its federal structure and devolved both policymaking and information dissemination to the state level. Even in that environment of heightened salience for state politics and issues of federalism, we find that existing frameworks still hold—partisanship still drives citizens’ trust in national leaders, particularly copartisans, for pandemic information. However, our findings also extend the existing literature in two ways. First, we find evidence that, at the state level, virus-mitigation policies broke down the relationship between partisanship and trust in information. In July 2022, as long as the state had mask mandates and stay-at-home orders in effect, Democrats had more trust in their state to provide them with reliable information about the pandemic regardless of the governor’s political party. At the state level, the power of partisan identity could be overcome by policies that addressed the threat posed by the pandemic. Second, we find that state identity matters in citizens’ choices of whom to trust for COVID-19 information, highlighting the need for consideration of national and state-level identities and attitudes in conjunction with one another. Most political-psychological frameworks that emphasize place-based identity focus on the role of national identity or the distinct experiences of rural citizens relative to urban ones. But in a political moment when federalism is highly salient, Americans rely on their attachment to their state to guide their assessment of information sources. State-level variables matter for and should be included in analyses of national political attitudes, particularly when federalism is highly salient. ## LITERATURE REVIEW Although the COVID-19 pandemic certainly changed day-to-day life and politics, it is reasonable to start with the assumption that Americans’ information habits in the face of this novel threat nonetheless can be understood through the lens of existing frameworks and theories in American political behavior. Four broad categories of literature can provide some guidance in answering the question of what drives citizens’ trust in their state and national officials to provide relevant pandemic information. First, a large existing literature identifies the features that tend to shape citizens’ trust in government and officeholders. As we try to better understand whom citizens turned to for information in a crisis, we begin by investigating the factors that shape citizens’ trust in their state and national officeholders generally. Second, literature on information-seeking behavior informs our understanding of whom citizens are most likely to turn to in moments of crisis, particularly when that crisis has been politicized. Third, we know that America’s pandemic response made significant use of the country’s federal structure and that citizens faced a federated information environment. Therefore, literature on citizens’ views of federalism and their preferences for policy decentralization is useful in understanding citizens’ relative trust in state and national leaders. That literature highlights the role of partisanship and ideology in those individual-level preferences—partisanship and ideology frequently shape the frames through which citizens analyze information from local, state, and national leaders. Finally, in a pandemic environment where geography and physical space matter a great deal, real differences in state and place-based identities and citizens’ attachments to their states may impact who they trust to provide pandemic-related information. Trust is a foundational component to democratic theory, as scholars in both American politics and political theory have observed. Trust is an individual-level evaluative or affective orientation that varies based on citizens’ understanding of the purposes of government (Miller 1974; Stokes 1962; Hetherington 2004). In a democratic society in particular, political trust is closely associated with political legitimacy and the rule of law and is needed to secure citizen compliance with the law (Barber 1983; Levi 1997, 1998; Scholz and Lubell 1998). When citizens trust their government and its officeholders, politicians are accorded greater political capital and institutional legitimacy (Easton 1967; Gamson 1968). In short, trust matters, perhaps especially in a life-and-death crisis situation such as the one the pandemic posed. Trust, particularly in a federal system where citizens are presented with choices of who to rely on for information and policy, is shaped by partisanship. Time and again, scholars have found that citizens prefer policymaking to be done by whichever level of government most closely aligns with their partisan affiliation (Wolak 2016; Dinan and Heckelman 2020; Riker 1964). The legal scholar Jessica Bulman-Pozen writes that “individuals’ beliefs about whether the state or federal government is the proper government to ‘run . . . things’ depend on which party is in control at both levels” (2014, 1120). Citizens have similarly shown a tendency to trust the national government more when it is controlled by copartisans (Morisi, Jost, and Singh 2019), and in general, people infer trustworthiness from partisan stereotypes, trusting copartisans over rival partisans (Carlin and Love 2013). Generally, states and state officials have enjoyed higher levels of political trust than the national government in recent years (Gallup 2020). Although national political trust has steadily declined since World War II, faith in state government has remained flat, and high (McCarthy 2018; Pew Research Center 2019). The scholarship on predictors of state trust falls into two camps. The first group of studies finds that citizens’ trust in their states generally follows from their trust in the national government or from their feelings about national conditions (Hetherington and Nugent 2001; Uslaner 2001). Others, however, argue that trust in specific institutions or leaders at the state level is driven by state economic performance or legislative professionalism (Kelleher and Wolak 2007; Richardson, Konisky, and Milyo 2012). In a more recent study, the political scientist Jennifer Wolak finds that citizens trust their state government more when it is controlled by copartisans, that trust in state governments does reflect state economic conditions, and that more homogeneous states engender more trust (Wolak 2020). The political scientists Bethany Albertson and Shana Kushner Gadarian’s (2015) research into information-seeking and trust in an anxious political environment provides additional insight into how Americans decide who to trust in a pandemic. They demonstrate that, in a public health emergency, people selectively trust specific experts (including the CDC, doctors, and friends in the medical field) over political figures, celebrities, and other government agencies (Albertson and Gadarian 2015). Anxiety also generates increased information-seeking behavior, and more biased patterns of information searching. Politicized threats will exacerbate those patterns (Albertson and Gadarian 2015). ## FEDERALISM MATTERS Given the decentralized nature of the pandemic response in the United States and the resulting salience of federalism, it would be reasonable to assume that citizens’ trust in their state leaders for pandemic information would be grounded in their personal ideological commitments and their views of federalism. If citizens hold strong beliefs about which level of government should engage in policy administration, they are likely to express trust in the officeholders at that level to provide pertinent information related to that policymaking. We might particularly expect this to be true in an environment when policy dispersion is a prominent topic. Generally, studies of public opinion and federalism have concluded that support for decentralized policymaking is highly correlated with partisanship and ideology (Schneider and Jacoby 2003; Konisky 2011; Thompson and Elling 1999). Particularly given that the COVID-19 pandemic highlighted American federalism, and at times direct competition between states over case rates and regulations, we can expect citizens’ ideological commitments in favor of or against decentralization to inform their information-seeking behavior. One aspect of the existing literature on attitudes toward state government that demands reevaluation in the context of the COVID-19 pandemic is the sense that issues of federalism and state authority are low salience and out of the public view. In the past, scholars argued that citizens face significant hurdles in collecting information on state government and state policymaking, and do not generally find state political action to be interesting or important (Delli Carpini and Keeter 1996; Jennings and Zeigler 1970). The extraordinarily heightened salience of state policy during the pandemic has almost certainly changed the landscape in terms of federalism and public opinion. Although recent literature has claimed that “intuitive federalism” allows citizens to make reasonable decisions about the allocation of policymaking authority, COVID policymaking almost certainly commands highly salient, explicit attitudes (Schneider and Jacoby 2003; Schneider, Jacoby, and Lewis 2011). Scholars of federalism have long wrestled with the question of whether American states reflect truly distinct identities. Myriad scholarship argues that federalism is only justified if states command loyalty from their citizens (Feeley and Rubin 2008; Riker 1964; Choudhry 2001). Most famously, political scientist Daniel Elazar (1966) argued that states did have unique cultures. More recently, scholars have asserted that state identity simply does not exist for most states (Feeley and Rubin 2008; Levy 2007). Place-based identity has been the subject of a surge in political psychology research in the past decade, but much of this work has focused on national identity or rural identity (see Cramer 2016; Jacobs and Munis 2019). Daniel Hopkins (2018), a political scientist, argues that state identities are also less political in nature; he finds that people report greater pride in their state’s landscapes, natural resources, or size than they do its political culture or values. Bulman-Pozen posits a partisan formulation of state identity that might be particularly helpful in understanding citizens’ relative trust in state and federal leaders to provide pandemic information. She argues that “our contemporary federal system generates a check on the federal government and fosters divided citizen loyalties . . . because it provides durable and robust scaffolding for partisan conflict” (2014, 1080–81). From this perspective, state-based identity exists, but it is fluid, partial, and based on the state’s ability to provide a competing vision of the national will when compared with the national government. If Bulman-Pozen is right, the pandemic should have provided a unique opportunity for states to offer alternative policy responses and alternative information environments to contrast with the national narrative. In the section that follows, we identify specific hypotheses about citizens’ trust in state and national leaders to provide pandemic information based on this literature. ## HYPOTHESES Hypotheses on citizen trust in leaders fall into three broad categories. First is the theory that trust in state leaders and institutions to provide reliable information is driven by partisanship. Second is whether ideology and ideological commitments to federalism drive higher trust in state leaders to provide pandemic-related information given the federated nature of the COVID-19 information environment. Third is that state identity matters and significantly influences how citizens seek out state-centered pandemic information at a moment of heightened salience for federalism. ### Partisanship The cited literature makes it clear that both citizens’ trust in officials and their information-seeking behavior is political. People tend to get their news and information from copartisan sources, and their trust in both state and national governments is often shaped by partisanship. As pandemic federalism took center stage in 2020, partisanship became the guiding narrative to explain state policy decision-making and citizens’ responses to those choices; poll after poll in 2020 demonstrated that regardless of state residency, relative to Democrats, Republicans were less likely to support policies such as mask mandates, stay-at-home orders and social distancing protocols, and were more hesitant to get a vaccine (Newport 2020a, 2020b, 2020c; Tyson 2020). The pandemic also took place during a highly contested, close, national election—a situation in which we might expect even normally nonpartisan events to become highly politicized. Even basic information was quickly subjected to politicization as Democrats and Republicans diverged in their understanding and analysis of how COVID-19 spread and its severity relative to other viruses. Previous work suggests that pandemic public health information would not be particularly politicized (Albertson and Gadarian 2015). the specifics of this pandemic, however, point in another direction. Elsewhere in this issue, the political scientist Sarah James and her colleagues discuss the highly polarized political environment that shaped responses to the pandemic. They point in particular to the ways in which partisan battles and an increasingly polarized political environment interfered with governors’ efforts to respond to the pandemic threat (James, Tervo, and Skocpol 2022, this issue). Given that we might ordinarily expect citizens’ trust and information preferences to be driven by partisanship, and that a looming presidential election and clashes between the president and governors over pandemic response measures were depicted in a highly partisan light, we expect that > H1a: Republicans will be more trusting in President Trump to provide pandemic information than Democrats. Although participants’ trust in Trump’s information will fall along partisan lines, we expect a different relationship between partisanship and trust in information coming from the CDC and Dr. Fauci, the leading U.S. infectious disease expert. A 2019 poll by the Pew Research Center finds that Democrats are more likely to support scientists playing an active role in policy debates, 54 percent (relative to 34 percent of Republicans) expressing a belief that scientific experts are better at making decisions about scientific issues than other people are (Funk et al. 2019). Studies that focus explicitly on Fauci’s expertise in the context of the COVID-19 pandemic suggest that his approval of policies such as vaccination can improve all partisan groups’ confidence and uptake of the policy—but that these effects are strongest for Democrats (Bokemper et al. 2021; Evans and Hargittai 2020). Given both the general findings about Democrats’ trust in experts and specific research about Fauci’s role in relaying pandemic information, we hypothesize that > H1b: Democrats will be more trusting than Republicans of the CDC and Dr. Fauci to provide pandemic information. National politics set the stage for a relationship between partisanship and trust in national leaders to provide pandemic-related information, but the president’s devolved policy approach also means that partisanship should affect trust in state leaders as information sources. In nearly every case, it was America’s governors who assumed leadership of the pandemic response (as opposed to state courts or legislatures) and thus gubernatorial partisanship likely became more prominent and salient for citizens who may have previously paid little attention to state-level partisan politics. The effects of partisanship on trust in information provided by state-level officials should mirror that at the national level—copartisans will produce more trustworthy information than members of the other party. However, public opinion research suggests not only that mass opinion is shaped by elite cues (Lenz 2012; Zaller 1992), but also that it can be particularly influenced by cues that appear to go against the party position (Chiang and Knight 2011). For example, the political scientist Guy Grossman and his colleagues (2020) find that, in the context of the COVID-19 response, state government leaders’ stay-at-home recommendations were more effective at reducing mobility in Democratic counties than Republican counties—what we would expect, given that Democrats were more anxious about the pandemic. Furthermore, the effect on Democrats’ mobility increased when stay-at-home recommendations were coming from Republican governors, because Republican governors who instituted strict COVID-19 policy were going against their national party’s stated preferences. In this case, seeing Republican governors act against their party’s position made Democrats even more likely to comply than when the policy was implemented by Democratic governors. Combined with the literature that suggests Americans are more likely to trust copartisans, we are left with two competing hypotheses: > H1c: Partisanship shapes trust in information from state officials, on the basis of the Governor’s party. > > H1d: Democrats are more likely to trust state officials to provide pandemic information if their state is governed by a Republican who supports more restrictive COVID-19 policy. ### Ideology Pandemic policymaking was highly devolved with regulations and policies coming from state, county, and sometimes even city officials, and citizens faced a highly federated set of information sources. We know that attitudes toward highly salient policy issues that evoke questions of federalism may be driven by citizens’ long-standing beliefs about political structure and the proper distribution of power in the American system (Green and Guth 1989). Many studies (such as Dinan and Heckelman 2020; Wolak 2016; Konisky 2011) find that attitudes toward issues of federalism are shaped by core ideological preferences for devolution and for decentralized policymaking. American conservatism has long embraced federalism and decentralized governance as a core belief. We therefore expect that > H2a: Conservatives’ trust in state officials is not shaped by the governor’s party. Although we expect conservatives to maintain a theoretical and ideological commitment to decentralized policymaking and information across the board, liberalism generally councils a preference for centralization and authority at the national level. The political scientist John Dinan and the economist Jae Heckelman (2020) find, as expected, that liberals are generally less supportive of decentralization. Essentially, this implies that federalism is salient for conservatives and not for liberals. Given that liberals have a general preference for centralization, but that partisanship is likely the stronger force in driving liberals’ attitudes toward state policymaking, we expect that > H2b: Liberals are more trusting of information from Dr. Fauci and the CDC than they are of state officials, even when state officials are copartisans. ### Identity If Elazar (1966) was correct that states have unique political cultures, and if those cultures remain clearly defined today, then they must have become extraordinarily salient during the COVID-19 pandemic. Time and again the news media highlighted distinct state demographics, geography, cultures, and politics as explanations for highly differentiated COVID case counts and lockdown policies. Partisanship and ideology are guiding explanations for Americans’ political attitudes, but a focus solely on these characteristics fails to capture the importance of place-based identity during the pandemic. We argue that the decentralized government handling of the pandemic also increased the salience of another identity—individuals’ attachment to their state and to the nation writ large. Research on national identity suggests that attachment to Americans as a group has a profound impact on political behavior from voting to one’s willingness to donate to charity (Huddy and Khatib 2007; Theiss-Morse 2009). Thus it seems equally plausible that those who feel most strongly connected to Americans as a group will also view federal-level information as particularly trustworthy in response to the pandemic. A shared sense of nationwide community and that we must protect all Americans would lead individuals to place their trust in the president, Congress, and other officials in Washington to provide pandemic-related guidance that would help the group. > H3a: Stronger national identity will increase trust in federal officials to provide pandemic information. Although the primacy of national over state identity might be the status quo in contemporary American politics, the COVID-19 pandemic and the subsequent response create a context in which attachment to home state and the sense of being a member of the state “group” influenced Americans’ attitudes toward government. Suddenly, state citizenship became a matter of life and death. State identity should therefore shape individuals’ relative trust in various leaders’ pandemic information. At the start of the lockdowns (or lack thereof) citizens who found themselves living in states to which they were closely attached were comforted, knowing that the pandemic response would be handled by trusted entities. Citizens who considered themselves strangers in a strange land, in contrast, were wary of state-based policymaking that might reflect precisely the political culture and distinct identity that they did not share. > H3b: Stronger state identity leads to greater trust in state officials to provide pandemic information. In summary, our hypotheses suggest that Americans’ trust in various government entities to provide reliable and accurate information is, to some extent, politics as usual. Partisanship and ideology will guide people’s feelings about the information they receive, with favorable attention to copartisans. Conservative commitments to devolution will lead them to be more trusting of information from lower levels of government. We recognize that neither of these expectations is novel in the context of American political behavior but find the possibility of evidentiary support for each to be reassuring—the discipline does not need to dismantle its frameworks for understanding behavioral phenomena in the face of a novel stimulus. However, we do see our final set of hypotheses, particularly H3b with its focus on state identity, as highlighting the disruptive potential of COVID-19. Given wide disparities in state pandemic response policies, and the high salience of state-based pandemic statistics such as case counts and death rates, it seems more likely that citizens’ trust in their state officials to handle COVID is related to their attachment to the state and sense of state-level community rather than a by-product of their trust in national leaders. If the pandemic heightened Americans’ awareness and understanding of federated policymaking and slowed the march toward a nationalized politics, political-psychological frameworks centered on national identities will need to focus more on individuals’ identification with their state and with state copartisans, as we do here. ## METHODOLOGY The survey used for this study drew on a representative sample of 1,0481 English-speaking Americans who were registered on the online participant-pool site Prolific (Palan and Schitter 2018) and was completed between July 24 and August 1, 2020 (for more information about how they collect representative samples, see Prolific Team 2019). Prolific users have been found to be reliable, honest participants in academic research who produce data quality comparable to MTurk and better than some other online research platforms (Peer et al. 2017), and their representative sampling schemes have been used for other research on COVID-19 attitudes (Oreffice and Quintana-Domeque 2020). Although we should be cautious about assuming generalizability from any nonprobability sample, this approach lets us consider the attitudes of a wide range of Americans at a financially affordable price. Even though Prolific’s representative sample stratifies across age, sex, and ethnicity to draw subgroups with the same proportions as the national population, it nonetheless is susceptible to the educational and partisan skew common to online samples in which participants are not drawn randomly from the population (see Berinsky, Huber, and Lenz 2012; Levay, Freese, and Druckman 2016). Thus, although our survey is representative of the United States on age, sex, and ethnicity,2 it oversamples the more highly educated at the expense of those without high school degrees and Democrats relative to Republicans and Independents (for a full demographic breakdown and comparison with census data, see table 1). To compensate for this sampling bias, we run our models in three ways: unweighted, with a correction for education, and with a correction for partisanship. Across the board, weighting on these single variables does not substantially alter the results of our analyses, reducing our concerns about the nonprobabilistic nature of our sample. We also considered running additional analyses with weights across multiple variables simultaneously. However, in some cases a very small number of observations (for example, seven participants who did not graduate from high school) are counting for a substantial amount of the weighted sample, blowing up the variance in the model and raising concerns about the introduction of additional bias (for concerns about bias introduced by weighting nonprobability surveys, see Kennedy et al. 2016). Because our goal is to reduce bias rather than introduce new ones, we see the single-variable weights as an effective methodological compromise. We include the models using the unweighted sample in the discussion and present the models with weights in appendix B; that our findings do not dramatically change when these weights are used increases our confidence in our findings. View this table: [Table 1.](http://www.rsfjournal.org/content/8/8/181/T1) **Table 1.** Characteristics of Sample Relative to the National Population To further reduce concerns about the quality of findings from our nonprobability sample at a single point in time, we ran additional analyses on publicly available data from the Axios-Ipsos Coronavirus Index, a series of probability-based surveys capturing Americans’ attitudes on COVID-19 and government handling of the pandemic. Although the questions in these surveys do not allow us to test all of our hypotheses, they give us insight into the first set of hypotheses (1a–1d) at two moments in time: July and October 2020. These analyses, included in appendix C, show similar relationships to those we find in our sample, increasing our confidence that the findings discussed are not unique to the individuals who chose to participate in our study; nor are they limited to the moment in time captured in our data. Participants who consented to participate in the survey were first asked what state they lived in and how long they had lived there, then answered a series of questions designed to measure their state and national identities. Both the national and state identity indices have been used in previous political science research (Huddy and Khatib 2007; Hopkins 2018) and reflect a larger set of psychological measures that tap subjective group identity (Huddy, Mason, and Aarøe 2015; Schildkraut 2011; Theiss-Morse 2009). Each index consists of four questions: “How important is being an [American/state demonym] to you?” “To what extent do you see yourself as a typical [American/state demonym]?” “How well does the term [American/state demonym] describe you?” and “When talking about [Americans/state demonym], how often do you say *we* instead of *they*?” Both sets of items have high internal consistency as measured by Cronbach’s alpha (state identity items’ alpha=0.91, national identity items; alpha=0.88) and were thus added together and standardized to create indices that run from zero (no identity with the state or nation) to one (highest possible identification with the state or nation). Although the distribution of national identity is slightly left-skewed (M = 0.63, median = 0.625, SD = 0.26), indicating that the average person identifies reasonably strongly as American, state identity is flatter and normally distributed (M = 0.5, median = 0.5, SD = 0.29). The average participant felt reasonably attached to their state identity, but some felt little identification with their state and others felt very strongly connected. Once they had completed the American and state identity scales, participants were asked a series of questions about their perceptions of the government’s handling of the pandemic. First was whether they saw the federal or state government as leading the response to the coronavirus outbreak in their area. The vast majority (79 percent) saw their state as the leader of the COVID-19 response; the remainder (21 percent) attributed the bulk of the response to the national government. They were then asked about their level of trust (using a 4-point scale that runs from not at all to a great deal) in various individuals and organizations to provide reliable information on coronavirus. This question explicitly operationalizes our primary outcome of interest—trust in information from government entities: how much do you trust the individuals and organizations below to provide reliable information on coronavirus? Participants evaluated a list of entities as part of this question, including President Trump and state government officials as well as the CDC, Dr. Anthony Fauci, and the World Health Organization (WHO). Table 2 displays the average trust in each entity across the entire sample; in general, participants placed the most trust in expert bureaucrats—Fauci, the CDC, and WHO—to provide reliable information, followed by state governments. Trump received the least trust across the full sample. Participants were also asked about their level of concern about coronavirus spread, and their general trust in the president, their state’s governor, and the federal and state governments. View this table: [Table 2.](http://www.rsfjournal.org/content/8/8/181/T2) **Table 2.** Average Trust in Information from Government Entities, Full Sample Respondents finished the survey by answering a series of demographic questions, including partisanship and ideology. Both partisanship and ideology were measured using the typical questions deployed by the American National Election Studies. Participants were first asked whether they usually thought of themselves as a Democrat, Republican, Independent, or something else, and then offered a follow-up question that allowed them to indicate the strength of their party identity or to note whether they lean more toward one major party or the other. To capture ideology, they were asked “which of the following best describes you?” and to choose a position on a 5-point scale from very liberal to very conservative. As mentioned and seen in table 1, the sample included more Democrats and liberals than the general population but still had a substantial number of Republicans and conservatives—more 20 percent. ## FINDINGS Although our hypotheses are broken down by partisanship, ideology, and identity, we start our discussion of findings in the context of attitudes toward the federal government versus those toward the state government. Hypotheses 1a, 1b, and 3a speak to trust in federal government entities and individuals to provide COVID-19 information, predicting that Republicans will be more trusting of information from Trump, Democrats of information from bureaucrats, and those who more strongly identify as Americans will be more trusting of federal entities’ information as a whole. Hypotheses 1c, 1d, 2a, 2b, and 3b all predict relationships between our three key variables of interest and trust in the state government to provide pandemic information. To examine the relationships between partisanship, national identity, and the federal response to COVID-19, we focus primarily on our three variables that capture trust in information from specific national-level individuals or agencies: President Trump, the CDC and Anthony Fauci. Each is included in a separate ordinary least squares (OLS) regression with the independent variables of interest, including partisanship, ideology, national and state identity, as well as control variables for the individual’s concern about the virus and demographic characteristics such as gender, education, race, and income. Table 3 presents results from the three OLS models. At first glance, we see support for hypotheses 1a and 1b and mixed support for 3a. Partisanship is one of the strongest predictors for trust in President Trump to provide information—and in the expected direction—with a move from the strongest partisans to partisan leaners equating to about half a point shift in trust. In other words, strong Republicans have almost a full point greater trust in Trump to provide pandemic information than strong Democrats. Partisanship is a statistically significant but substantively weaker explanation for trust in the federal bureaucracy’s information, producing at most a half-point shift in trust in information from Fauci between strong Democrats and Republicans and an even smaller change for trust in the CDC to provide information on COVID. As expected, we see that the directionality of the relationship between partisanship and trust flips; whereas Republicans are more trusting of Trump to provide pandemic information, Democrats are more trusting of Fauci and the CDC. National identity plays a statistically significant but weaker role in shaping trust toward both Trump and the CDC and has a nonsignificant effect on trust in Fauci for pandemic information. Moving from one end of the national identity scale—those who do not identify at all as Americans—to the other—those whose American identity is central to their self-conception, only increases trust in Trump and the CDC by 0.03 to 0.05 on a 4-point scale. Although we predicted that national identity would shape trust in Trump, the CDC and Fauci to provide information on the crisis, it is possible that participants were making a distinction between the president and a bureaucratic agency as national decisionmakers relative to Fauci as an individual expert. View this table: [Table 3.](http://www.rsfjournal.org/content/8/8/181/T3) **Table 3.** Effects of Key Variables on Trust in Information from Federal COVID-19 Responders The most powerful predictor of trust in each federal entity to provide information was individual concern about the coronavirus; those who were more concerned were less likely to trust the president and more likely to trust Fauci and the CDC. The effect of coronavirus concern on trust in the pre+sident was equivalent to the effect of partisanship, but played a much stronger role in shaping trust in the CDC and Fauci to guide individuals’ understanding of the pandemic—those who were not at all concerned about the virus were a full point less trusting in all three entities than their very concerned peers. Turning to trust in the state government’s provision of information, our hypotheses depend on the partisan affiliation of each state’s governor and their policy response to the virus. We added three variables to the survey data to capture this variability. First, we included a dummy variable that took on a value of one if the participant lived in a state with a Republican governor. Sixty percent of our sample lived in states with Democratic governors. Second, we included a measure of whether a participant’s state had a mask mandate as of July 24, 2020 (CDC 2021). Seventy-five percent of participants lived in states with mask mandates as of late July 2020. Finally, we created a combined measure that took on a value of zero if the state had no mask policy, one if a Democratic governor had implemented the mask policy, and two if a Republican governor had instated a mask mandate.3 These variables help us assess hypotheses 1c, 1d, 2a, and 2b. Hypotheses 1b and 1c present slightly different takes on the relationship between participant partisanship and the party affiliation of their state’s governor. Hypothesis 1b suggests that partisans trust copartisans—Democrats will express greater trust in their state to provide pandemic information when they live in states with Democratic governors at the helm, and Republicans will feel the same when a Republican leads their state. Hypothesis 1c argues that a Republican governor breaking with the national narrative on partisan attitudes toward masking can actually receive even greater trust from Democrats than copartisan governors. Using OLS regression models, we once again find partial evidence for both hypotheses (see models in the appendix). Figure 1 displays the linear prediction of trust in the state government to provide pandemic information for Democrats, Republicans, and true Independents (leaners were categorized with partisans), broken down by whether they are represented by a Democratic or Republican governor. Although we see no significant differences in Republicans’ trust in the state government across gubernatorial partisan affiliation, we do see a half-point increase in Independents’ trust when they are governed by a Democrat, and a full point increase in Democrats’ trust under copartisans, both of which are statistically significant. In more substantive terms, this suggests that Democrats trust their state governments to provide information “not much” when they have a Republican at the helm, and Democrat-led states “a fair amount.” Meanwhile, Republicans trust their state government somewhere between “not much” and “a fair amount” regardless of the partisan affiliation of the state executive. ![Figure 1.](http://www.rsfjournal.org/https://www.rsfjournal.org/content/rsfjss/8/8/181/F1.medium.gif) [Figure 1.](http://www.rsfjournal.org/content/8/8/181/F1) **Figure 1.** Partisan Trust in Information from State Governments *Source*: Authors’ calculations. *Note*s: Predictions based on an OLS regression (full model available in appendix). Trust in state government is measured on a 4-point scale, 0 indicating no trust and 3 indicating a great deal of trust. Participants who identified as independents who lean toward a party were categorized as partisans. Figure 2 shows us that when assessing hypothesis 1c we see a similar pattern for Republicans. Their trust in the state government’s information does not depend on the partisan affiliation of the governor or the implementation of a mask mandate in their state. However, in states where Republican governors have acted contrary to national partisan narratives about support for masking, we do see increased trust in state government information among Independents and Democrats. These differences are not statistically different from one another for Independents but are for Democrats. Although trust in information from the state government under mask-mandating Republican governors does not surpass that under a copartisan Democratic governor, thereby failing to support our hypothesis, the improved trust in states where these governors have instituted mask mandates offers some support for the argument that contra-party positional cues can be particularly important for members of the opposite party. ![Figure 2.](http://www.rsfjournal.org/https://www.rsfjournal.org/content/rsfjss/8/8/181/F2.medium.gif) [Figure 2.](http://www.rsfjournal.org/content/8/8/181/F2) **Figure 2.** Democrats’ Trust in Information from State Governments Based in Gubernatorial Party and Policy *Source*: Authors’ calculations. *Note*s: Predictions based on an OLS regression (full model available in appendix). Trust in state government measured on a 4-point scale, zero indicating no trust and three indicating a great deal of trust. Participants who identified as independents who lean toward a party were categorized as partisans. The consistent pattern in the relationship between respondents’ party identity and their governor’s party affiliation also plays out when we shift to looking at ideology, per hypotheses 2a and 2b. In support of these two hypotheses, figure 3 displays results from an OLS regression that includes the interactive effect of participant ideology and gubernatorial partisanship on trust in the state government to provide pandemic guidance. From the figure, it is easy to see that conservatives’ trust in the state government is not shaped by state party control, whereas liberals, much like Democrats, see a full point increase in their trust in state government information when the state is run by a Democrat.4 ![Figure 3.](http://www.rsfjournal.org/https://www.rsfjournal.org/content/rsfjss/8/8/181/F3.medium.gif) [Figure 3.](http://www.rsfjournal.org/content/8/8/181/F3) **Figure 3.** Liberals Trust Information from State Governments More Under Democratic Governors *Source*: Authors’ calculations. *Note*s: Predictions based on an OLS regression (full model in appendix A). Trust in state government is measured on a 4-point scale, 0 indicating no trust and 3 indicating a great deal of trust. We can also take a more descriptive approach to assessing hypotheses 2a and 2b. Although our measures do not allow for a direct comparison of individuals’ trust in national government information relative to state government information (the ideal for assessing liberals’ commitment to centralized policy response), we can nonetheless assess the average trust each group has in the information from federal and state officials and compare those values. Because we ask about specific government officials at the national level, but about the more generic state government, we collapse participants’ trust in Trump, Fauci, and the CDC into a general trust in information from the federal government measure, and create a second trust in information from federal bureaucrats measure that does not include Trump. As expected, the two indices look very similar for conservatives (trust in bureaucrats = 1.69, trust in federal government = 1.69), but less so for liberals (trust in bureaucrats = 2.33, trust in government=1.65). Finally, our last hypothesis (3b) speaks to the role of state identity—attachment to state community and culture—in participants’ trust in the state to provide pandemic information. To assess the role of state identity, we return to a simpler OLS model of the relationship between trust in state government and our list of key independent variables and controls without any interactive effects, similar to those used in table 1 to explain trust in federal entities. The results depicted in figure 4 to facilitate comparisons of the variables’ substantive effects. We see support for our hypothesis: as state identity increases, the increase in trust in state government information is small but significant. Substantively, this translates to a shift of about 0.06 on the trust scale when moving from someone who feels no connection to their state to someone who expresses extreme state identity. Although state identity has nowhere near the explanatory power of gubernatorial party affiliation, it nonetheless plays an important role in shaping state trust, as it remains statistically significant across the range of interactive models as well (see appendix). ![Figure 4.](http://www.rsfjournal.org/https://www.rsfjournal.org/content/rsfjss/8/8/181/F4.medium.gif) [Figure 4.](http://www.rsfjournal.org/content/8/8/181/F4) **Figure 4.** Relative Change in Trust in Information from the State Government Across Key Variables *Source*: Authors’ calculations. *Note*: Points represent coefficients from an OLS regression (full model available in appendix A), with 95 percent confidence intervals. Variables in bold have a statistically significant effect on trust in state government to handle the COVID-19 pandemic. In summary, we find some measure of support for each theory—partisanship, ideology, and state and national identity all affect Americans’ trust in different levels of government to provide pandemic information. Republicans were more likely to trust the president, a copartisan, to effectively provide information, and Democrats to trust federal bureaucrats and their state governments, especially when those states were run by Democratic governors. Although Republican governors’ implementation of mask mandates increased Democrats’ trust in their state governments relative to governors who did not implement similar policies, it was not enough to overcome the copartisan preference. In contrast, Republican and conservative trust in their state governments held fast across gubernatorial party affiliation and pandemic policy implementation, potentially reflecting an underlying commitment to devolution and decentralization, or, as the political scientist Deborah Schildkraut and her colleagues argue, that “at this point in time, the central philosophical difference between liberals and conservatives over the scope of government seems to have given way to partisanship” (2020). Finally, our results show that place-based identities also play a role in shaping individuals’ trust in various levels of government to provide pandemic information. Whereas those who more strongly identify as Americans were more trusting of both Trump and the CDC to guide their understanding of the pandemic, strong state identifiers were more likely to trust their state government. ## DISCUSSION In the highly salient and highly federated information environment created by the COVID-19 pandemic, partisanship, ideology, and place-based identity all affected whom citizens chose to trust for crucial pandemic-related information. Elsewhere in this issue, scholars show that trust in government and in officeholders became, during the pandemic, a life and death issue. As Page-Tan and her colleagues demonstrate, communities with access to information through horizontal and vertical linkages were more likely to engage in behaviors designed to slow the spread of the coronavirus (Page-Tan, Marion, and Aldrich 2022, this issue). Suhay and her colleagues show that individuals’ trust in government was similarly correlated with engaging in protective health behaviors (Suhay et al. 2022, this issue). In each case, who citizens chose to trust for pandemic-related information ultimately led either to behavior associated with better health outcomes or greater risk for those individuals and their communities The story of political trust during the COVID-19 pandemic is obviously a partisan one. Across many of the articles in this issue, research shows just how much polarization impacted policy decisions and individual behavioral choices. Our research adds to that narrative by demonstrating that individual-level partisanship was associated with greater trust in copartisans to provide pandemic-related information. In this sense, the COVID-19 pandemic was, perhaps surprisingly, a kind of politics as usual. Existing, dominant frameworks in political science that point to partisanship as the driving force behind political information-seeking behavior held true, even in the context of a public health emergency when we might have expected politics to fade into the background. Early in the pandemic, it seemed as though the highly federated nature of pandemic policymaking and information dissemination might prove to be a critical juncture, turning Americans’ attention away from the national politics, narratives, and information that had come to dominate their lives. COVID-19 did, perhaps, prove to be a disruptive point in the general trend toward nationalized policymaking—pandemic policies varied remarkably from one state to the next. But citizens still turned to national copartisans for trustworthy information. Our results also show, however, that other individual-level identities, including place-based and ideological attachments, influenced citizens information-seeking behavior in a time of heightened information salience, and a particularly federated information environment. What is more, policies designed to counter the threats posed by the pandemic dramatically reduced the power of partisanship at the state level. Democrats in our study did not care whether their governor was a Republican—if he or she had overseen the implementation of mask mandates and stay-at-home orders, they trusted the state government to provide them with reliable information. These findings suggest that existing political science frameworks focused on national partisan narratives might be helpfully augmented by more focus on state-level partisanship and state identities. Our research is limited in important and significant ways. We do not, for instance, hypothesize about other potentially important individual-level factors that likely influence pandemic trust behavior including, for example, a citizen’s relative concern about the coronavirus. Nor do we demonstrate a link between trust in information from different sources and behavioral changes that could promote or prevent community transmission. However, an increasing body of research suggests that we should expect a link between individuals’ trust in governments to provide information and their willingness to adhere to certain types of behavior, both generally and in the specific context of the COVID-19 pandemic (Marien and Hooghe 2011; Pagliaro et al. 2021; Suhay et al. 2022, this issue; Zhao et al. 2020). As Page-Tan and her colleagues (2022, this issue) suggest, vertical ties—government officials, mayors, and other elected representatives—are important guides in adopting health practices. Our survey also reflects a particular moment in time. The summer of 2020 presented citizens with a set of federated information choices that were different than those available at other points in the pandemic. Suhay and colleagues find that between March and November 2020, trust in officials at every level of government declined, but that the decline was uneven (Suhay et al. 2022, this issue). We have no reason to believe that the factors we point to (partisanship, ideology, and place-based identity) became unimportant over that period, and our supplemental analyses of Axios-Ipsos poll data from July and October 2020 suggest that, at least for a short time, the patterns we have identified hold. We are cautious to say anything about how citizens’ relative weighing of their various identities may have changed over the extended course of the pandemic as their need for information and the available sources of that information also shifted. Finally, our findings are limited by the nonprobabilistic nature of our sampling process. Although we have run a variety of robustness checks to feel confident about the validity of the analyses above, we nonetheless acknowledge that our methodological approach biases our findings. For example, although our sample distribution of Democrats and Republicans mirrors the national numbers as reported by Gallup, almost 40 percent of our sample are strong partisan identifiers. We might expect strong identifiers to be guided by partisanship more than the average American, thus overestimating the actual effect of partisanship on our dependent variables of interest. We also have a much higher proportion of highly educated participants than is present in the national population. These individuals are more knowledgeable about the pandemic and more worried about the effects of COVID-19 (Rattay et al. 2021). Thus they might be more responsive to policy implementation, leading to the overestimation of the power of mask mandates to overcome shared partisan identity as a motivator for trust in information from the state government. In spite of these concerns, our auxiliary analyses lend credence to our findings and we appreciate the flexibility offered when fielding an original survey, particularly the ability to ask participants about their state identity and more fully integrate questions about federal and state-level attitudes. The political scientists Malcom Feeley and Edward Rubin (2008) argue that federalist structures should only be used when citizens’ identities are truly divided and that federalism is only truly justified by the existence of varied and divided place-based identities among the citizenry. In other words, America’s devolution of pandemic policymaking to the states should have been justified by its citizens’ state attachments. As we broadly assess the lasting impact of COVID-19 on America’s federal system, it will be crucial to understand how citizens’ identities—whether partisan, ideological, or place-based—shaped their views of state and national responses. In states where citizens already identified strongly with their state governments, either because they shared the partisan identity of the state leaders or they felt a strong sense of belonging to their state community, the pandemic afforded an opportunity to reinforce the necessity of the federalist system. In places where people only weakly identified with their states or disagreed with the partisan policy preferences of their leaders, the nation’s decentralized pandemic response may well bolster calls for reform. James, Tervo, and Skocpol (2022, this issue) argue that, from a policy response perspective, federalism failed to provide Americans with the policymaking tools required to manage the COVID-19 pandemic. In fact, they argue, our reliance on federalism for the pandemic response left Americans vulnerable to policymaking failures. Given the obvious centrality of federalism in the story of the COVID-19 pandemic, future research will need to account for both the ways that our federated response may have failed in the execution of acute public health policies, and the ways that a federated information system and opportunities for local, state, and federal leadership may have engendered the trust of citizens where a unitary state would have failed. ## APPENDIX A: OLS MODELS UNDERLYING FIGURES 1–4 View this table: [Table A.1.](http://www.rsfjournal.org/content/8/8/181/T4) **Table A.1.** Effect of Partisanship, Gubernatorial Party Affiliation, and Gubernatorial Mask Mandates on Trust in Information from State Government View this table: [Table A.2.](http://www.rsfjournal.org/content/8/8/181/T5) **Table A.2.** Effect of Ideology and Gubernatorial Party Affiliation on Trust in State Government Information View this table: [Table A.3.](http://www.rsfjournal.org/content/8/8/181/T6) **Table A.3.** Effects of Key Variables on Trust in Information from State COVID-19 Responders ## APPENDIX B: OLS MODELS USING CORRECTIVE PARTISAN WEIGHTS We ran each of our models three ways: unweighted, weighted by the census distribution of the U.S. population on education, and weighted on Gallup’s distribution of the U.S. population’s partisan identity. We used single weights rather than any sort of combination of factors to avoid introducing greater bias into our results; with so few people in certain demographic and political categories, we were concerned about increasing the variance in certain categories. This appendix displays the unweighted and weighted models for each test next to one another for ease of comparison. The vast majority of key relationships stay the same regardless of the weighting choices. Differences that do exist do not substantially affect the specific hypotheses. They do highlight one pattern that impacts our conclusions: the relationship between party identification and national identity. Several models suggest that national identity plays a statistically significant role in relevant measures of trust in information when we weight on party identification but is statistically insignificant in the unweighted and education-weight models. In each of those cases, national identity positively affects trust in information from state officials. Thus this article could underestimate the impact of national identity on trust in information from state officials because our survey undersamples independents and Republicans relative to the national population. Research based on the Grinnell College National Poll has found a positive relationship between national identity and support for Donald Trump (Rawhouser-Mylet and Hanson 2020); our data also suggest that this is the case. We find that national identity and partisanship (measured on the 7-point scale that combines identity and strength) are correlated at 0.36, such that strong Republicans are more likely to also score highly on the national identity scale. The correlation is stronger (0.48) if we look at approval of President Trump. However, these moderate correlations should not be taken to suggest that national identity and partisanship are the same underlying concept. Instead, they highlight the importance of controlling for partisanship when making claims about the effects of national identity, and for the possibility that national identity operates differently for Democrats and Republicans. View this table: [Table B.1.](http://www.rsfjournal.org/content/8/8/181/T7) **Table B.1.** Trust in Information from President Trump Across Weighting Schemes View this table: [Table B.2.](http://www.rsfjournal.org/content/8/8/181/T8) **Table B.2.** Trust in Information from the CDC Across Weighting Schemes View this table: [Table B.3. ](http://www.rsfjournal.org/content/8/8/181/T9) **Table B.3.** Trust in Information from Dr. Fauci Across Weighting Schemes View this table: [Table B.4. ](http://www.rsfjournal.org/content/8/8/181/T10) **Table B.4.** Trust in Information from State Officials Across Weighting Schemes View this table: [Table B.5. ](http://www.rsfjournal.org/content/8/8/181/T11) **Table B.5.** Effect of Partisanship and Gubernatorial Party Affiliation on Trust in State Government Information Across Weighting Schemes View this table: [Table B.6. ](http://www.rsfjournal.org/content/8/8/181/T12) **Table B.6.** Effect of Partisanship, Gubernatorial Party Affiliation, and Mask Mandates Across Weighting Schemes View this table: [Table B.7. ](http://www.rsfjournal.org/content/8/8/181/T13) **Table B.7.** Effect of Partisanship, Gubernatorial Party Affiliation, and Stay-at-Home Orders on Trust in Information from State Officials Across Weighting Schemes View this table: [Table B.8. ](http://www.rsfjournal.org/content/8/8/181/T14) **Table B.8.** Effect of Ideology and Gubernatorial Party Affiliation on Trust in State Government Information ## APPENDIX C: FINDINGS USING ALTERNATIVE DATA SOURCES Our data are from an online sample of participants who self-selected into the study, but the findings parallel those from other studies in the summer of 2020. Axios-Ipsos, a national polling collaboration, conducted a series of surveys about Americans’ attitudes toward the coronavirus and the government handling of the pandemic, beginning in March 2020. The seventeeth wave of the survey was fielded at the time nearest to our own data collection, running from July 17 through 20, 2020, and collecting responses from 1,037 individuals through the web-enabled KnowledgePanel, the oldest and largest probability-based online panel in the United States (Ipsos 2022). To examine whether our results are unique to July 2020, we also run the same analyses on an additional wave of the survey conducted in October 2020 (sample size 1,079). The Axios-Ipsos surveys do not allow us to exactly duplicate our analyses because they do not include all the measures we incorporated into our survey. Most notably, they ask participants about their ideology or their state identity, which means we are unable to evaluate hypotheses 2a, 2b, 3a, and 3b using these datasets. Their value, however, comes from their measure of partisan identity and of trust in government information, which is phrased similarly to that used in our survey. In both the July and October waves, participants were asked “How much trust do you have in each of the following to provide you with accurate information about coronavirus or COVID-19?” In July, they were asked to consider five entities: the federal government, their state government, the Centers for Disease Control and Prevention (CDC), national public health officials, and the White House. In October, participants were asked about their trust in the federal government, their state government, the CDC, national public health officials, Donald Trump, and Joe Biden. Thus we can effectively assess the effect of partisan identity on trust in information from President Trump (H1a), information from the CDC (H1b), and state officials (H1c and H1d). Hypothesis 1a argues that Republicans would be more trusting in President Trump to provide pandemic information than Democrats would. The October 2020 survey also allows us to examine the converse of this relationship: the expectation that Democrats will be more trusting of information from (at the time) Democratic presidential candidate Joe Biden. Table C.1 demonstrates that this is the case and that, if anything, the role of partisanship in predicting trust in information from national-level elected officials (or candidates) only increased as we approached the November 2020 presidential election. As is clear from column 1, Republican trust in information from the White House was a full point higher than that of Democrats in July. By October, Republican trust in information from Trump was 1.4 points higher that of Democrats, but trust in Democratic presidential candidate Joe Biden was 1.3 points lower. Thus the data from the Axios-Ipsos polls offers further support for hypothesis 1a and highlights the way in which partisan identification shaped partisan trust in information not only from President Trump but also from Democratic candidate Joe Biden. We also see support for our hypothesis that Democrats will be more trusting federal bureaucrats (H1b), particularly the CDC and national public health officials. Across both waves of the Axios-Ipsos poll, Republicans are about 0.14 to 0.3 points less trusting in information from the CDC and national public officials than their Democratic counterparts (see table C.2). This partisan effect is substantially smaller in size than that on trust in Trump and Biden to provide reliable information but nonetheless suggests that individuals were assessing information from nonpartisan bureaucrats differently than they were elected officials and candidates for federal office. What is more, the effect sizes for the relationship between partisanship and trust in both partisan elected officials and nonpartisan public health officials in the Axios-Ipsos poll waves are similar to those found in our sample. Although the results for hypotheses 1a and 1b reflect those found in our Prolific sample, those for hypotheses 1c retain the same pattern of findings but with weaker effect sizes. Hypothesis 1c focuses on the interaction between participants’ partisan identification and the party affiliation of their state’s governor, whereas 1d expects trust in information from the state government to depend on the interaction between participants’ partisanship and the state’s decision to implement strict COVID-19 containment policies. Figure C.1 depicts the relationship between participant partisanship, gubernatorial party affiliation, and trust in information from the state government, replicating the analysis of our own data shown in figure 3 (full regression models in table C.3). Whereas Democrats in our study were a full point more trusting of information from the state government if it was led by a copartisan, the Axios-Ipsos participants’ trust in information from the state government during the same period shows no statistically significant difference. In the October data, we see the expected effect, Democrats trusting information from state governments run by a copartisan about half a point more than those run by Republicans. Turning to Republican participants, we see little difference in our data in Republican trust in information under Republican or Democratic governors; in contrast, in both July and October, the Republicans in the Axios-Ipsos poll trust information from state governments more when the governor is a copartisan. In short, although the effects are weaker in some contexts, the Axios-Ipsos data still reinforces our finding that partisanship shapes trust in information from the state government and gives us added leverage in understanding Republican attitudes. View this table: [Table C.1. ](http://www.rsfjournal.org/content/8/8/181/T15) **Table C.1.** Trust in Information from Federal Elected Officials, July and October 2020 View this table: [Table C.2.](http://www.rsfjournal.org/content/8/8/181/T16) **Table C.2.** Trust in Information from Federal Bureaucrats Finally, the Axios-Ipsos data reinforce our argument that policy implementation designed to address the threat posed by the pandemic reduced the impact of partisanship on trust in information from the state government, particularly for Democrats. In July 2020, Democrats trust in information from state governments run by Republican or Democratic governors are statistically indistinguishable, as long as they have implemented a mask policy. In October, partisanship has regained some of its power—Democrats have higher trust in information from governments run by copartisans who have implemented mask mandates than they do if their state is run by the out-party—but the presence of a mask mandate still strongly affects trust in information from the state. We also see in this data that Republicans are not making a distinction across gubernatorial party or policy; if anything, they are more trusting of information from the state government when that state has not implemented a mask mandate. Ultimately, the data from the two Axios-Ipsos waves reinforce our findings about the role of partisanship in shaping responses to COVID-19, particularly the trust Americans had in information from state and federal leaders. By examining this data at two different time points, we can feel more confident that the relationships we identify in our original survey are not an artifact of the sample or the period at which the data is collected, but an enduring pattern of behavior. ![Figure C.1.](http://www.rsfjournal.org/https://www.rsfjournal.org/content/rsfjss/8/8/181/F5.medium.gif) [Figure C.1.](http://www.rsfjournal.org/content/8/8/181/F5) **Figure C.1.** Trust in Information from the State Government, by Partisanship and Governor’s Partisan Identification *Source*: Authors’ calculations based on data from Axios-Ipsos Coronavirus Index, waves 17 and 29. *Note*s: Predictions based on an OLS regression (full model in table C.3). Trust in state government is measured on a 4-point scale, 0 indicating no trust and 3 indicating a great deal of trust. ![Figure C.2.](http://www.rsfjournal.org/https://www.rsfjournal.org/content/rsfjss/8/8/181/F6.medium.gif) [Figure C.2.](http://www.rsfjournal.org/content/8/8/181/F6) **Figure C.2.** Trust in Information from State Government, by Partisanship, Governor’s Partisan Affiliation, and Policy Implementation *Source*: Authors’ calculations based on data from Axios-Ipsos Coronavirus Index, waves 17 and 29. *Note*s: Predictions based on an OLS regression (full model in table C.3). Trust in state government is measured on a 4-point scale, 0 indicating no trust and 3 indicating a great deal of trust. View this table: [Table C.3.](http://www.rsfjournal.org/content/8/8/181/T17) **Table C.3.** Trust in Information from State Government, Considering Elite Partisanship and Policy ## FOOTNOTES * 1. An a priori power analysis for logistic regressions with an odds ratio of 1.3 and alpha equal to 0.05 calculates the required sample size at 988, calculated using G*Power (Faul et al. 2009). * 2. The categories used on Prolific to prescreen for ethnicity are White, Mixed, Asian, Black, and Other, so although our sample is representative of the national population when it comes to white and black participants, it underrepresents Hispanic identification. * 3. Eight states had Republican governors who instituted mask mandates by the end of July: Alabama, Arkansas, Indiana, Maryland, Massachusetts, Ohio, Texas, and West Virginia. Only one state with a Democratic governor—Wisconsin—had not implemented a mask mandate. We also ran these tests using stay-at-home orders and see the same patterns for stay-at-home orders that we do mask mandates. For length purposes, our results relating to stay-at-home orders are included in appendix B, table B.7. * 4. Although participants’ identification on the ideology and partisan scales are highly correlated (r = 0.72 when using the 5- and 7-point scales that assess strength of identification), approximately 35 percent of participants hold ideological positions that are different from those associated with the contemporary Democratic and Republican Parties. * © 2022 Russell Sage Foundation. Pears, Emily, and Emily Sydnor. 2022. “COVID-19 and the Culture of American Federalism.” *RSF: The Russell Sage Foundation Journal of the Social Sciences* 8(8): 181–220. DOI: 10.7758/RSF .2022.8.8.09. Special thanks to Southwestern University and the Sam Taylor Fellowship for funding the survey described in this article. Direct correspondence to: Emily Pears, at Emily.Pears{at}claremontmckenna.edu, 850 Columbia Ave., Claremont, CA 91711, United States; Emily Sydnor, at sydnore{at}southwestern.edu, 1001 E. University Ave., Georgetown, TX 78626, United States. Open Access Policy: *RSF: The Russell Sage Foundation Journal of the Social Sciences* is an open access journal. This article is published under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License. ## REFERENCES 1. Albertson, Bethany, and Shana Kushner Gadarian. 2015. *Anxious Politics: Democratic Citizenship in a Threatening World*. New York: Cambridge University Press. 2. Barber, Bernard. 1983. *The Logic and Limits of Trust*. New Brunswick, N.J.: Rutgers University Press. 3. Berinsky, Adam J., Gregory A. Huber, and Gabriel S. Lenz. 2012. “Evaluating Online Labor Markets for Experimental Research: Amazon.com’s Mechanical Turk.” Political Analysis 20(3): 351–68. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1093/pan/mpr057&link_type=DOI) 4. Bokemper, Scott E., Gregory A. Huber, Alan S. Gerber, Erin K. James, and Saad B. Omer. 2021. “Timing of COVID-19 Vaccine Approval and Endorsement by Public Figures.” Vaccine 39(5): 825–29. 5. Bulman-Pozen, Jessica. 2014. “Partisan Federalism.” Harvard Law Review 127(4): 1077–45. 6. Campbell, Angus, Philip E. Converse, Warren E. Miller, and Donald E. Stokes. 1960. *The American Voter*. Chicago: University of Chicago Press. 7. Carlin, Ryan E., and Gregory J. Love. 2013. “The Politics of Interpersonal Trust and Reciprocity: An Experimental Approach.” Political Behavior 35: 43–63. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1007/s11109-011-9181-x&link_type=DOI) 8. Centers for Disease Control and Prevention (CDC). 2021. “National Environmental Public Health Tracking Network: U.S. State and Territorial Orders Requiring Masks in Public.” Accessed January 10, 2021. [https://ephtracking.cdc.gov/DataExplorer/](https://ephtracking.cdc.gov/DataExplorer/). 9. Chiang, Chun-Fang, and Brian Knight. 2011. “Media Bias and Influence: Evidence from Newspaper Endorsements.” Review of Economic Studies 78(3): 795–820. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1093/restud/rdq037&link_type=DOI) [Web of Science](http://www.rsfjournal.org/lookup/external-ref?access_num=000292765400001&link_type=ISI) 10. 1. Kalypso Nicolaidis and 2. Robert Howse Choudhry, Sujit. 2001. “Citizenship and Federations: Some Preliminary Reflections.” In *The Federal Vision: Legitimacy and Levels of Governance in the United States and the European Union*, edited by Kalypso Nicolaidis and Robert Howse. London: Oxford University Press. 11. Cramer, Katherine J. 2016. *The Politics of Resentment*. Chicago: University of Chicago Press. 12. Delli Carpini, Michael X., and Scott Keeter. 1996. *What Americans Know about Politics and Why It Matters*. New Haven, Conn.: Yale University Press. 13. Dinan, John, and Jac C. Heckelman. 2020. “Stability and Contingency in Federalism Preferences.” Public Administration Review 80(2): 234–43. 14. Easton, David. 1967. *A Systems Analysis of Political Life*. New York: John Wiley & Sons. 15. Elazar, Daniel J. 1966. *American Federalism: A View from the States*. New York: Thomas Y. Crowell. 16. Evans, John H., and Eszter Hargittai. 2020. “Who Doesn’t Trust Fauci? The Public’s Belief in the Expertise and Shared Values of Scientists in the COVID-19 Pandemic.” Socius: Sociological Research for a Dynamic World 6. First published online August 6, 2020. DOI: [https://doi.org/10.1177/2378023120947337](https://doi.org/10.1177/2378023120947337). 17. Faul, Franz, Edgar Erdfelder, Axel Buchner, and Albert-Georg Lang. 2009. “Statistical Power Analyses Using G*Power 3.1: Tests for Correlation and Regression Analyses.” Behavior Research Methods 41(4): 1149–60. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.3758/BRM.41.4.1149&link_type=DOI) [PubMed](http://www.rsfjournal.org/lookup/external-ref?access_num=19897823&link_type=MED&atom=%2Frsfjss%2F8%2F8%2F181.atom) [Web of Science](http://www.rsfjournal.org/lookup/external-ref?access_num=000271930000018&link_type=ISI) 18. Feeley, Malcolm M., and Edward Rubin. 2008. *Federalism: Political Identity and Tragic Compromise*. Ann Arbor: University of Michigan Press. 19. Funk, Cary, Meg Hefferon, Brian Kennedy and Courtney Johnson. 2019. “Trust and Mistrust in Americans’ Views of Scientific Experts.” Washington, D.C.: Pew Research Center. Accessed July 23, 2022. [https://www.pewresearch.org/science/2019/08/02/partisanship-influences-views-on-the-role-and-value-of-scientific-experts-in-policy-debates/](https://www.pewresearch.org/science/2019/08/02/partisanship-influences-views-on-the-role-and-value-of-scientific-experts-in-policy-debates/). 20. Gallup. 2020. “Trust in Government.” Accessed July 23, 2022. [https://news.gallup.com/poll/5392/Trust-Government.aspx](https://news.gallup.com/poll/5392/Trust-Government.aspx). 21. Gamson, William A. 1968. *Power and Discontent*. Homewood, Ill.: Dorsey Press. 22. Green, Donald P., Bradley Pamquist, and Eric Schickler. 2002. *Partisan Hearts and Minds*. New Haven, Conn.: Yale University Press. 23. Green, John C., and James L. Guth. 1989. “The Missing Link: Political Activists and Support for School Prayer.” Public Opinion Quarterly 53(1): 41–57. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1086/269140&link_type=DOI) 24. Grossman, Guy, Soojong Kim, Jonah M. Rexer, and Harsha Thirumurthy. 2020. “Political Partisanship Influences Behavioral Responses to Governors’ Recommendations for COVID-19 Prevention in the United States.” Proceedings of the National Academy of Sciences 117(39): 24144–53. [Abstract/FREE Full Text](http://www.rsfjournal.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoicG5hcyI7czo1OiJyZXNpZCI7czoxMjoiMTE3LzM5LzI0MTQ0IjtzOjQ6ImF0b20iO3M6MjA6Ii9yc2Zqc3MvOC84LzE4MS5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 25. Hetherington, Marc J. 2004. *Why Trust Matters: Declining Political Trust and the Demise of American Liberalism*. Princeton, N.J.: Princeton University Press. 26. 1. John R. Hibbing, 2. Elizabeth Theiss-Morse, and 3. James H. Kuklinski Hetherington, Marc J., and John D. Nugent. 2001. “Explaining Public Support for Devolution: The Role of Political Trust.” In *What Is It about Government That Americans Dislike?*, edited by John R. Hibbing, Elizabeth Theiss-Morse, and James H. Kuklinski. New York: Cambridge University Press. 27. Hopkins, Daniel J. 2018. *The Increasingly United States: How and Why American Political Behavior Nationalized*. Chicago: University of Chicago Press. 28. Huddy, Leonie, and Nadia Khatib. 2007. “American Patriotism, National Identity, and Political Involvement.” American Journal of Political Science 51(1): 63–77. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1111/j.1540-5907.2007.00237.x&link_type=DOI) [Web of Science](http://www.rsfjournal.org/lookup/external-ref?access_num=000243229900005&link_type=ISI) 29. Huddy, Leonie, Lilliana Mason, and Lene Aarøe. 2015. “Expressive Partisanship: Campaign Involvement, Political Emotion, and Partisan Identity.” American Political Science Review 109(1): 1–17. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1017/S0003055414000604&link_type=DOI) 30. Ipsos. 2022. “KnowledgePanel.” Accessed October 25, 2022. [https://www.ipsos.com/en-us/solutions/public-affairs/knowledgepanel](https://www.ipsos.com/en-us/solutions/public-affairs/knowledgepanel). 31. Jacobs, Nicholas F., and B. Kal Munis. 2019. “Place-Based Imagery and Voter Evaluations: Experimental Evidence on the Politics of Place.” Political Research Quarterly 72(2): 263–77. 32. James, Sarah, Caroline Tervo, and Theda Skocpol. 2022. “Institutional Capacities, Partisan Divisions, and Federal Tensions in U.S. Responses to the COVID-19 Pandemic.” RSF: The Russell Sage Foundation Journal of the Social Sciences 8(8): 154–80. DOI: [https://doi.org/10.7758/RSF.2022.8.8.08](https://doi.org/10.7758/RSF.2022.8.8.08). 33. Jennings, M. Kent, and Harmon Zeigler. 1970. “The Salience of American State Politics.” American Political Science Review 64(2): 523–35. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.2307/1953847&link_type=DOI) [Web of Science](http://www.rsfjournal.org/lookup/external-ref?access_num=A1970G722800010&link_type=ISI) 34. Jones, Jeffrey M. 2021. “Quarterly Gap in Party Affiliation Largest Since 2012.” Gallup Inc., April 7. Accessed October 25, 2022. [https://news.gallup.com/poll/343976/quarterly-gap-party-affiliation-largest-2012.aspx](https://news.gallup.com/poll/343976/quarterly-gap-party-affiliation-largest-2012.aspx). 35. Kelleher, Christine A., and Jennifer Wolak. 2007. “Explaining Public Confidence in the Branches of State Government.” Political Research Quarterly 60(4): 707–21. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1177/1065912907304496&link_type=DOI) [Web of Science](http://www.rsfjournal.org/lookup/external-ref?access_num=000251205300011&link_type=ISI) 36. Kennedy, Courtney, Andrew Mercer Scott Keeter, Nick Hatley, Kyley McGeeney, and Alejandra Gimenez. 2016. “Evaluating Online Nonprobability Surveys.” Washington, D.C.: Pew Research Center. Accessed July 23, 2022. [https://www.pewresearch.org/methods/2016/05/02/evaluating-online-nonprobability-surveys/](https://www.pewresearch.org/methods/2016/05/02/evaluating-online-nonprobability-surveys/). 37. Konisky, David. 2011. “Public Preferences for Environmental Policy Responsibility.” Publius: The Journal of Federalism 41(1): 76–100. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1093/publius/pjp044&link_type=DOI) 38. Lenz, Gabriel S. 2012. *Follow the Leader?: How Voters Respond to Politicians’ Policies and Performance*. Chicago: University of Chicago Press. 39. Levay, Kevin E., Jeremy Freese, and James N. Druckman. 2016. “The Demographic and Political Composition of Mechanical Turk Samples.” SAGE Open 6(1). First published March 15, 2016. DOI: [https://doi.org/10.1177/2158244016636433](https://doi.org/10.1177/2158244016636433). 40. Levi, Margaret. 1997. *Consent, Dissent, and Patriotism*. New York: Cambridge University Press. 41. 1. Karen S. Cook and 2. Russell Hardin, edited by 3. Valerie Braithwaite and 4. Margaret Levi Levi, Margaret. 1998. “A State of Trust.” In *Trust and Governance*, by Karen S. Cook and Russell Hardin, edited by Valerie Braithwaite and Margaret Levi. New York: Russell Sage Foundation. 42. Levy, Jacob T. 2007. “Federalism, Liberalism, and the Separation of Loyalties.” American Political Science Review 101(3): 459–77. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1017/S0003055407070268&link_type=DOI) [Web of Science](http://www.rsfjournal.org/lookup/external-ref?access_num=000248817900006&link_type=ISI) 43. Marien, Sophie, and Marc Hooghe. 2011. “Does Political Trust Matter? An Empirical Investigation into the Relation Between Political Trust and Support for Law Compliance.” European Journal of Political Research 50(2): 267–91. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1111/j.1475-6765.2010.01930.x&link_type=DOI) [Web of Science](http://www.rsfjournal.org/lookup/external-ref?access_num=000286051900006&link_type=ISI) 44. McCarthy, Justin. 2018. “Americans Still More Trusting of Local Than State Government.” Gallup, October 8. Accessed July 23, 2022. [https://news.gallup.com/poll/243563/americans-trusting-local-state-government.aspx](https://news.gallup.com/poll/243563/americans-trusting-local-state-government.aspx). 45. Miller, Arthur H. 1974. “Rejoinder to ‘Comment’ by Jack Citrin: Political Discontent or Ritualism?” American Political Science Review 68(3): 989–1001. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.2307/1959142&link_type=DOI) [Web of Science](http://www.rsfjournal.org/lookup/external-ref?access_num=A1974U237200003&link_type=ISI) 46. Morisi, Davide, John T. Jost, and Vishal Singh. 2019. “An Asymmetrical President-in-Power Effect.” American Political Science Review 113(2): 614–20. 47. Newport, Frank. 2020a. “The Partisan Gap in Views of the Coronavirus.” Gallup, May 15. Accessed July 23, 2022. [https://news.gallup.com/opinion/polling-matters/311087/partisan-gap-views-coronavirus.aspx](https://news.gallup.com/opinion/polling-matters/311087/partisan-gap-views-coronavirus.aspx). 48. Newport, Frank. 2020b. “Further Dissecting the Partisan Gap in Views of the Coronavirus.” Gallup, July 31. Accessed July 23, 2022. [https://news.gallup.com/opinion/polling-matters/316520/further-dissecting-partisan-gap-views-coronavirus.aspx](https://news.gallup.com/opinion/polling-matters/316520/further-dissecting-partisan-gap-views-coronavirus.aspx). 49. Newport, Frank. 2020c. “Partisanship and Vaccine Uptake Strategies.” Gallup, December 18. Accessed July 23, 2022. [https://news.gallup.com/opinion/polling-matters/328034/partisanship-vaccine-uptake-strategies.aspx](https://news.gallup.com/opinion/polling-matters/328034/partisanship-vaccine-uptake-strategies.aspx). 50. Oreffice, Sonia, and Climent Quintana-Domeque. 2020. “Gender Inequality in Covid-19 Times: Evidence from UK Prolific Participants.” *IZA* discussion paper no. 3648803. Bonn: IZA Institute of Labor Economics. Accessed July 23, 2022. [https://covid-19.iza.org/publications/dp13463/](https://covid-19.iza.org/publications/dp13463/). 51. Page-Tan, Courtney, Summer Marion, and Daniel P. Aldrich. 2022. “Information Trust Falls: The Role of Social Networks and Information During the COVID-19 Pandemic Among Suburbanites.” RSF: The Russell Sage Foundation Journal of the Social Sciences 8(8): 32–51. DOI: [https://doi.org/10.7758/RSF.2022.8.8.02](https://doi.org/10.7758/RSF.2022.8.8.02). 52. Pagliaro Stefano, Simona Sacchi, Maria Giuseppina Pacilli, Marco Brambilla, Francesca Lionetti, Karim Bettache, et al. 2021. “Trust Predicts COVID-19 Prescribed and Discretionary Behavioral Intentions in 23 Countries.” PLoS ONE 16(3): e0248334. DOI: [https://doi.org/10.1371/journal.pone.0248334](https://doi.org/10.1371/journal.pone.0248334). [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1371/journal.pone.0248334&link_type=DOI) 53. Palan, Stefan, and Christian Schitter. 2018. “Prolific. Ac—A Subject Pool for Online Experiments.” Journal of Behavioral and Experimental Finance 17 (March): 22–27. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1016/j.jbef.2017.12.004&link_type=DOI) 54. Peer, Eyal, Laura Brandimarte, Sonam Samat, and Alessandro Acquisti. 2017. “Beyond the Turk: Alternative Platforms for Crowdsourcing Behavioral Research.” Journal of Experimental Social Psychology 70 (May): 153–63. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1016/j.jesp.2017.01.006&link_type=DOI) 55. Pew Research Center. 2019. “Public Trust in Government: 1958–2019.” Pew Research Center: U.S. Politics & Policy, April 11. Accessed July 23, 2022. [https://www.pewresearch.org/politics/2019/04/11/public-trust-in-government-1958-2019/](https://www.pewresearch.org/politics/2019/04/11/public-trust-in-government-1958-2019/). 56. Prolific Team. 2019. “Representative Samples.” Prolific, May 31. Accessed July 23, 2022. [https://researcher-help.prolific.co/hc/en-gb/articles/360019236753](https://researcher-help.prolific.co/hc/en-gb/articles/360019236753). 57. Rattay, Petra, Niels Michalski, Olga Maria Domanska, Anna Kaltwasser, Freia De Bock, Lothar H. Wieler, and Susanne Jordan. 2021. “Differences in Risk Perception, Knowledge, and Protective Behaviour Regarding COVID-19 by Education Level Among Women and Men in Germany. Results from the COVID-19 Snapshot Monitoring (COSMO) Study.” PLoS One 16(5): e0251694. DOI: [https://doi.org/10.1371/journal.pone.0251694](https://doi.org/10.1371/journal.pone.0251694). [PubMed](http://www.rsfjournal.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Frsfjss%2F8%2F8%2F181.atom) 58. Roawhouser-Mylet, Georgia, and Peter Hanson. 2020. “American National Identity and Attitudes Towards President Trump.” Data Analysis and Social Inquiry Lab at Grinnell College, May 8. Accessed October 25, 2022. [https://dasil.sites.grinnell.edu/2020/05/american-national-identity-and-attitudes-towards-president-trump/](https://dasil.sites.grinnell.edu/2020/05/american-national-identity-and-attitudes-towards-president-trump/). 59. Richardson, Lilliard E., David M. Konisky, and Jeffrey Milyo. 2012. “Public Approval of State Legislatures.” State Politics and Policy Quarterly 37(1): 99–116. 60. Riker, William H. 1964. *Federalism: Origin, Operation, Significance*. New York: Little, Brown. 61. Saad, Lydia. 2021. “Americans’ Political Ideology Held Steady in 2020.” Gallup Inc., January 11. Accessed October 25, 2022. [https://news.gallup.com/poll/328367/americans-political-ideology-held-steady-2020.aspx](https://news.gallup.com/poll/328367/americans-political-ideology-held-steady-2020.aspx). 62. 1. Seth J. Schwartz, 2. Koen Luyckx, and 3. Vivian L. Vignoles Schildkraut, Deborah J. 2011. “National Identity in the United States,” In *Handbook of Identity Theory and Research*, edited by Seth J. Schwartz, Koen Luyckx, and Vivian L. Vignoles. New York: Springer. 63. Schildkraut, Deborah J., Jeffrey M. Berry, and James M. Glaser. 2020. “Public Opinion About Federalism During the Coronavirus Pandemic.” Data for Progress (*blog*), May 4. Accessed July 23, 2022. [https://www.dataforprogress.org/blog/5/4/public-opinion-federalism](https://www.dataforprogress.org/blog/5/4/public-opinion-federalism). 64. Schneider, Saundra K., and William G. Jacoby. 2003. “Public Attitudes Toward the Policy Responsibilities of the National and State Governments: Evidence from South Carolina.” State Politics & Policy Quarterly 3(3): 246–69. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1177/153244000300300302&link_type=DOI) 65. Schneider, Saundra K., William G. Jacoby, and Daniel C. Lewis. 2011. “Public Opinion Toward Intergovernmental Policy Responsibilities.” Publius: The Journal of Federalism 41(1): 1–30. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1093/publius/pjq036&link_type=DOI) 66. Scholz, John T., and Mark Lubell. 1998. “Trust and Taxpaying: Testing the Heuristic Approach to Collective Action.” American Journal of Political Science 42(2): 398–417. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.2307/2991764&link_type=DOI) [Web of Science](http://www.rsfjournal.org/lookup/external-ref?access_num=000072446200003&link_type=ISI) 67. 1. Harlan Cleveland and 2. Harold D. Lasswell Stokes, Donald E. 1962. “Popular Evaluations of Government: An Empirical Assessment.” In *Ethics and Bigness*, edited by Harlan Cleveland and Harold D. Lasswell. New York: Klaus Reprint Company. 68. Suhay, Elizabeth, Aparna Soni, Claudia Persico, and Dave E. Marcotte. 2022. “Americans’ Trust in Government and Health Behaviors During the COVID-19 Pandemic.” RSF: The Russell Sage Foundation Journal of the Social Sciences 8(8): 221–44. DOI: [https://doi.org/10.7758/RSF.2022.8.8.10](https://doi.org/10.7758/RSF.2022.8.8.10). 69. Theiss-Morse, Elizabeth. 2009. *Who Counts as an American?: The Boundaries of National Identity*. Cambridge: Cambridge University Press. 70. Thompson, Lyke, and Richard Elling. 1999. “Let Them Eat Marblecake: The Preferences of Michigan Citizens for Devolution and Intergovernmental Service-Provision.” Publius: The Journal of Federalism 29(1): 139–54. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1093/oxfordjournals.pubjof.a030005&link_type=DOI) 71. Tyson, Alec. 2020. “Republicans Remain Far Less Likely Than Democrats to View COVID-19 as a Major Threat to Public Health.” *FactTank* (Pew Research Center blog), July 22. Accessed July 23, 2022. [https://www.pewresearch.org/fact-tank/2020/07/22/republicans-remain-far-less-likely-than-democrats-to-view-covid-19-as-a-major-threat-to-public-health/](https://www.pewresearch.org/fact-tank/2020/07/22/republicans-remain-far-less-likely-than-democrats-to-view-covid-19-as-a-major-threat-to-public-health/). 72. 1. John R. Hibbing, 2. Elizabeth Theiss-Morse, and 3. James H. Kuklinski Uslaner, Eric. 2001. “Is Washington Really the Problem?” In *What Is It about Government That Americans Dislike?*, edited by John R. Hibbing, Elizabeth Theiss-Morse, and James H. Kuklinski. New York: Cambridge University Press. 73. Wolak, Jennifer. 2016. “Core Values and Partisan Thinking about Devolution.” Publius: The Journal of Federalism 46(4): 463–85. [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1093/publius/pjw021&link_type=DOI) 74. Wolak, Jennifer. 2020. “Why Do People Trust Their State Government?” State Politics & Policy Quarterly 20(3): 313–29. 75. Zaller, John R. 1992. *The Nature and Origins of Mass Opinion*. New York: Cambridge University Press. 76. Zhao, Erfei, Qiaoa Wu, Eileen M. Crimmins, and Jennifer A. Ailshire. 2020. “Media Trust and Infection Mitigating Behaviors During the COVID-19 Pandemic in the USA.” BMJ Global Health 5:e003323. [Abstract/FREE Full Text](http://www.rsfjournal.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NToiYm1qZ2giO3M6NToicmVzaWQiO3M6MTI6IjUvMTAvZTAwMzMyMyI7czo0OiJhdG9tIjtzOjIwOiIvcnNmanNzLzgvOC8xODEuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9)