Local Impact of Global Crises,Institutional Trust,and Consumer Well-Being: Evidence from the COVID-19 Pandemic

Global Crises and Their Local Impact

Global crises are defined as “crises whose origins and outcomes cannot for the most part be confined inside the borders of particular nation states; rather, they are endemic to, enmeshed within, and potentially encompassing of today’s late-modern, capitalistic … world (Cottle 2011, p. 78; emphasis in the original). Although global crises are usually approached through an economic perspective as “periodic disturbances in the world economy” (Burke 1988), the term has been expanded to capture global challenges related to the environment (e.g., air pollution, global warming), geopolitics (e.g., wars, terrorism), public health (e.g., diseases, malnutrition), and social issues (e.g., discrimination, income inequality) (Gill 2011). Growing evidence suggests that global crises have increased in frequency, severity, and spatial impact and that they cannot be resolved locally (Biggs et al. 2011). Within the last decade, the world has experienced several shocks that could be characterized as global crises, including the U.S. mortgage crisis in 2008, the ISIS terrorist attacks in 2015, the refugee crisis following the Syrian Civil War, environmental disasters caused by climate change (e.g., Australian wildfires, South Asia floods), and, more recently, the COVID-19 pandemic.

All these ostensibly unrelated events share two common features. First, they have a “deterritorializing impact” that spans across national borders even when their origin can be accurately pinpointed (Cottle 2009, p. 497). Second, they are facilitated by the same forces that underlie globalization’s advancement (e.g., the free movement of people, advances in technology and travel, emergence of global media, integration of national economies; Ritzer 2007). Despite these commonalities, the footprint of global crises across different geographical areas is neither quantitatively nor qualitatively homogeneous. For example, the financial crisis of 2008 had a stronger impact on advanced economies’ gross domestic product growth than on that of emerging economies (Berkmen et al. 2012), and the political success of populist parties in Europe that emerged as a by-product of the sovereign debt crisis was much stronger in European countries that participated in the Eurozone than in those that did not (Guiso et al. 2019). Conversely, climate change has a more severe distributional impact on poor countries than on richer ones (Mendelsohn, Dinar, and Williams 2006). Even the COVID-19 pandemic, which has affected the whole world, has led to varying infection and mortality rates across countries (Chaudhry et al. 2020).

Multiple factors could explain cross-country variability in global crises’ impact (e.g., national social structures, macroeconomic policies, geo-environmental factors, demographic composition)—which are beyond the scope of this research. Global crises are experienced in a unique, localized fashion within the areas they impact, and public acceptance and/or effectiveness of institutional interventions aimed at eliminating crises, mitigating their damaging consequences, and restoring consumer well-being must consider the interplay between a crisis’s global reach and its localized manifestation in the context of a specific spatial unit of analysis (e.g., community, region, country).

Trust in National and International Institutions During a Global Crisis

The onset of a crisis engenders the question of which institutions are responsible for regulating it. Claiming a central role in managing crises—and subsequently the right to exercise authority within the context of a crisis—crisis-regulating institutions axiomatically become subjects of public scrutiny in terms of both their political legitimacy and their effectiveness in preventing/handling the crisis (Easton 1979). Crisis-handling institutions are assessed against two types of legitimacy: normative legitimacy, or the objective evaluation of an institution’s legal right to exercise authority, and descriptive legitimacy, or the subjective assessment of the effectiveness of an institution’s actions in achieving positive outcomes (Beetham 1991). The two forms of legitimacy do not always go hand in hand. Institutions may enjoy normative legitimacy yet lack descriptive legitimacy (e.g., the “Not My President” rallies against the Trump administration) or hold descriptive legitimacy while lacking normative legitimacy (e.g., nongovernmental organizations supporting public-opinion-friendly causes without having the legal authority to enforce their views). As descriptive legitimacy is critical to achieving public acceptance of (and compliance with) an institution’s directives (Caldeira and Gibson 1992), we follow a descriptive approach in assessing the role of crisis-regulating institutions.

For locally confined crises, local institutions are seen as both responsible (in terms of political legitimacy) and able (in terms of knowledge and resources) to manage the crisis. Global crises, though, require regulatory efforts from supranational institutions that exceed a country’s geographical borders, need “institutionalized modes of social coordination” that extend beyond a nation’s sovereign authority, and demand nonhierarchical collaboration with institutions outside the nation-state (Findlay 2013, p. 6). The outcome of this coordination is essential for building institutional trust, the social capital of legitimacy (Torgler 2008).

The construct of trust holds a central place in social and political sciences. Trust “involves an individual making herself vulnerable to another individual, group, or institution that has the capacity to do her harm or betray her” (Levi and Stoker 2000, p. 476). Although there is individual-level variation in people’s willingness to trust other entities, trust is a relational construct addressed to a particular referent (Williams 2020). Depending on that referent, social scientists have studied different forms of trust, such as social trust (i.e., trust people place in other members of their social environment) and political trust (i.e., trust placed in specific political entities) (Newton, Stolle, and Zmerli 2018). In this research, we focus on institutional trust, a particular form of political trust that refers to crisis-regulating institutions (be they national or international). Different forms of trust relate to each other. Social trust correlates with institutional trust; people who are more trustful of other people tend to trust institutions more easily (Zmerli, Newton, and Montero 2007). However, as institutions are perceived as organizations responsible for delivering specific outcomes, institutional trust is also a function of an institution’s ability to fulfill citizens’ expectations (Warren 1999).

Empirical evidence suggests that global crises with a pronounced local impact cause distrust in both national and international institutions. Public opinion surveys measuring trust in the EU and national governments showed a steep decline in trust in the EU during the European debt crisis (from 60% in 2004 to 36% in 2015), mostly attributable to debtor states such as Greece, Spain, and Portugal, which experienced crisis’s impact more severely than other EU states. In contrast, the creditor countries in northern Europe that were less affected by the debt crisis retained high levels of trust in the EU during the crisis. In addition to decreased trust in the EU, citizens of countries most strongly hit by the crisis reported very low confidence in their national governments, unlike creditor countries, which retained stable levels of national government confidence throughout the crisis (Foster and Frieden 2017). In the context of the 2015 immigration crisis, data from the annual Eurobarometer indicate that in Germany and Greece—two of the countries that accepted disproportionately higher numbers of refugees than fellow EU states—EU trust indices (Germany: −23%; Greece: −11%) and national government trust indices (Germany: −12%; Greece: −21%) fell substantially more than the corresponding European index averages, reaching their lowest values in a decade (Eurobarometer 2015). Thus, even when a crisis exceeds strict national borders, the more severely affected local areas are more likely to distrust the national and international institutions they hold responsible for (not) preventing and/or (mis)handling the crisis.

Crisis Attributions as Explanatory Mechanism of Institutional Blame and (Dis)trust

Attribution theory investigates how individuals use information to derive causal explanations for events (Fiske and Taylor 1991). Causal attributions determine the role a person’s actions play in an outcome (Kelley and Michela 1980; Weiner 1980). As an entity is held more responsible for an outcome in which their behavior could have made a difference, causal attributions explain responses to unexpected negative situations (Weiner 1980).

Individuals make causal attributions on the basis of the locus, stability, and controllability of negative events (Folkes 1988; Russell 1982). The locus dimension reflects individuals’ beliefs about whether the cause of the event is internal or external, the stability dimension reflects individuals’ beliefs about whether the cause of the event is temporary or permanent, and the controllability dimension reflects individuals’ beliefs about whether the cause could have been changed or affected (Russell 1982; Weiner 1980). According to the culpable control perspective of blame attributions, when individuals review negative events, they evaluate the source’s causal influence over the event, assess whether the event was caused intentionally, and consider whether the event could have been prevented (Alicke et al. 2008). These evaluations determine perceptions of how much control the source had over the event.

The theory of blame posits that individuals assign blame when they (1) detect an event that violates a norm and (2) determine that an agent caused the norm-violating event (Malle, Guglielmo, and Monroe 2014). If no agent can be causally linked to the event, blame cannot emerge, as there is no target at which to direct the blame. If an agent is causally linked to the event, individuals evaluate whether the agent caused the event intentionally. If the agent is judged as having acted intentionally, individuals consider the agent’s reasons for behaving the way they did, and the resulting blame depends on how justifiable the agent’s behavior was. If the agent is judged as acting unintentionally, individuals consider the agent’s controllability over the event—that is, whether the agent should have (obligation) and could have (capacity) prevented the event (Malle, Guglielmo, and Monroe 2014). People blame agents more for violations the agents could control than for violations they could not (Darley and Schultz 1990).

In our context, the negative event is not the global crisis per se (i.e., the pandemic) but the local impact of the pandemic in the consumer’s own country. Applying attribution theory and the theory of blame, we argue that causal attributions of blame for the local impact of the global crisis based on locus or stability are difficult. Regarding locus, as the impact of the global crisis is experienced by individuals locally, the causes leading to that particular impact at the local level will inevitably be attributed internally and will be more severe for countries facing a disproportionately higher impact from the crisis compared with less affected areas. However, even when the cause of high local impact is situated internally, individuals may still base locus attributions not only on internal factors (e.g., incompetent national institutions) but also on situational factors (e.g., new COVID-19 variants). Moreover, in the case of global crises, it is not possible to attribute the locus of the heightened local impact only internally while totally discounting the role of external actors involved in the crisis (e.g., foreign countries, international institutions). For example, in the case of the pandemic, many people still attribute the severe impact of the pandemic on their country to external agents such as foreign countries (e.g., China, where the pandemic started) or international institutions (e.g., the World Health Organization [WHO], for not urging countries to contain the pandemic earlier).

Causal attributions of blame for the local impact of a global crisis based on stability are also unlikely because the trajectory of an ongoing global crisis is unknown and depends on more than just local actions. The primary consideration of stability-based attributions is whether the cause of the failure remains stable or fluctuates over time—that is, whether the causes of high local impact are temporary or permanent (Folkes 1988). Such assessments are also difficult because national and international crisis-regulating institutions adjust their policies across time to address the pandemic’s unpredictable development over time.

Consequently, causal attributions of blame for the local impact of a global crisis are mostly made on the basis of controllability. Assuming that high local impact was not volitional on the part of either national or international institutions, the primary consideration in assigning causal attributions of blame based on controllability is whether the local impact of the global crisis could be controlled, prevented, or mitigated in a superior way by those responsible to manage it (i.e., institutions). As global crises are events with deterritorialized reach and a non–locally specified origin (Cottle 2011), individuals perceive international institutions as agents having both the obligation and the capacity to address the crisis (or, in political theory terms, as institutions with descriptive legitimacy to act). In parallel, though, in countries where the local impact of the crisis is disproportionately high, similar judgments should be made for national institutions because citizens are expected to attribute cross-country differences in crisis impact on the mismanagement of the crisis at the local level. Thus, consumers in countries experiencing high local impact from a global crisis will blame both international and national institutions more (compared with low-local impact-countries).

Causal attributions of controllability for a negative event damage trustworthiness (Tomlinson and Mayer 2009). Thus, if individuals conclude that the cause of the crisis is attributed to (national) international institutions’ inability to control the crisis, blame toward (national) international institutions will emerge, and subsequently, trust in them will erode.

H₁: Perceptions of high relative local impact following a global crisis have a negative effect on (a) international institutional trust, mediated via international institutional blame, and (b) national institutional trust, mediated via national institutional blame.

Congruence Versus Compensation: Variation in Institutional Blame and Distrust

Sociopolitical trust literature proposes two opposing principles regarding how people form trust toward national and international institutions. The congruence principle posits that trust in national and international institutions go in tandem: as trust in the former increases (or decreases), so does trust in the latter (Muñoz, Torcal, and Bonet 2011). In a global crisis context, this prediction relies on the premise that global crises require the collaboration of national and international institutions; thus, blame and/or trust emanating from management of the crisis is directed in similar ways at both. This principle assumes that both local and global institutions are legitimized to handle a global crisis and proposes dual expectations of institutional conduct, against which institutional performance is compared. The compensation principle alternatively posits that trust in national and international institutions move in opposite directions because people see international institutions either as trustworthy counterweights to undependable national institutions or as unnecessary sovereignty threats to efficient national institutions—but not both (Muñoz, Torcal, and Bonet 2011). According to this principle, the distribution of institutional trust is a zero-sum game, whereby citizens should either assign the weight for crisis-management to international institutions by marginalizing the role of national institutions or confide in national institutions at the expense of their international counterparts who are deemed unauthorized, pragmatically unable, or morally illegitimate to prevent, manage, or mitigate the consequences of the crisis.

We argue that whether consumers will follow the congruence or the compensation principle when assessing national and international institutional interventions during global crises depends on two individual-level traits: their attitude toward globalization and their political ideology. We predict that individuals with positive globalization attitudes (because of their global affinity and legitimization of transnational governance) will follow the congruence principle (and thus retain high levels of trust for both national and international institutions during crisis), while political conservatives (because of their affinity for national sovereignty and punitive morality) will follow the compensation principle (and thus transfer blame and distrust from national to international institutions).

The Congruence Principle and the Role of Globalization Attitudes

Globalization attitude is defined as “support or opposition to globalization based upon the individual’s belief about the economic consequences of globalization” (Spears, Parker, and McDonald 2004, p. 58) or “the degree to which globalization is perceived positively, with the benefits to the local economy exceeding the demands placed on the local economy” (Suh and Smith 2008, p. 132). Although these definitions tap into beliefs about the consequences of globalization in the economic domain, people holding positive globalization attitudes also have favorable views about globalization as a social force or a cultural phenomenon. Individuals can be categorized as “globalists” (i.e., people seeing globalization as a force for good) or “antiglobalists” (i.e., people perceiving globalization as an economic or social threat) (Dimofte, Johansson, and Ronkainen 2008). Globalists tend to more positively view entities designated as international (e.g., brands, firms) than antiglobalists (Riefler 2012).

Positive globalization attitudes promote support for international political institutions (Furia 2005). People with a cosmopolitan political orientation—a correlate of globalization attitude—tend to view supranational organizations (e.g., the European Union, the United Nations, the North America Free Trade Agreement) positively (Norris 2000). Beliefs about the legitimacy of international institutions explain these favorable views. Positive globalization attitudes correlate with strong support for the internationalization of political authority on the grounds of functional and moral interdependence (Ecker-Ehrhardt 2012). Functional interdependence captures beliefs that international institutions are more effective than national ones in solving transnational issues, while moral interdependence reflects beliefs that international governance is ethically responsible for addressing injustices inextricably linked to globalization. In contrast, people with negative globalization attitudes view international institutions as illegitimate forms of governance that exacerbate the democratic deficit brought about by citizens’ inability to influence international institutions or hold them accountable through democratic processes (Machida 2009).

Globalization attitudes are not static; people adjust their beliefs about the desirability of international institutions experientially (Schaffer and Spilker 2016). Because globalization forces exacerbate the impact of global crises on local communities, globalists propound that international institutions should pay their moral duty by stepping up and exercising the authority handed to them to effectively handle the crisis locally (Ecker-Ehrhardt 2012). Essentially, even when blame is fairly assigned to international institutions for their inability to fully prevent or mitigate a crisis globally (i.e., even when obligation and capacity assessments render international institutions as credible targets of blame), globalists should limit such blame and retain high levels of trust in international institutions as actors who are both morally and functionally capable of dealing with a crisis not only at the global level but also in areas most hit by it. In line with the congruence principle, though, the globalists’ affinity for international institutions does not result in subsequent marginalization of national institutions. Although globalists might not view national institutions as morally responsible for the root causes of a global crisis, they perceive them as both ethically and pragmatically legitimized to manage the crisis’ local symptoms. Counter to views suggesting that globalists promote global governance at the expense of national sovereignty, we suggest that globalists advocate collaborative crisis management, view national and international institutions as “communicating vessels,” and trust both for crisis resolution (Ecker-Ehrhardt 2012).

H₂: Positive globalization attitudes (a) decrease international institutional blame and (b) increase national and international institutional trust.

The Compensation Principle and the Role of Political Ideology

Global crises constitute political events. It is well-established that liberals are more likely to believe that the global climate crisis is real and act against it (McCright et al. 2016). Left-wing ideology correlates with behaviors supporting human rights and predicts tolerance to military action combating international human rights violations (Cohrs et al. 2007). Conservative beliefs increased among older and disadvantaged citizens of New Zealand in the aftermath of the 2008 global financial crisis (Milojev et al. 2015), while the European sovereign debt crisis caused increased support for populist parties across the continent (Guiso et al. 2019). Thus, political orientation matters in the context of global crises.

In line with the compensation principle, we expect that, unlike globalists, political conservatives will view international and national institutions as conflicting entities such that blaming/distrusting the former coincides with absolving/trusting the latter. We argue that conservatives treat national and international institutions as competing crisis-regulating agents rather than collaborative players with aligned interests. This perspective ultimately works in favor of national institutions because conservatives show not only strong national attachment and animosity toward transnational governance but also more punitive and less trusting attitudes (thus legitimizing national and delegitimizing international institutions).

Political conservatism places high value on punishment as an appropriate response to transgressions, mostly because of the shared moral foundations underlying political and religious conservatism (Silver and Silver 2017). Political conservatism is associated with an increased tendency to punish wrongdoers across different life domains. For example, conservative parents are more likely to punish their children using corporal punishment (Ellison, Bartikowski, and Segal 1996), and people with conservative political orientation report higher levels of disgust toward perceived violations of moral purity issues (e.g., they are more likely to hold prejudicial attitudes toward LGBTQ+ people; Terrizzi, Shook, and Ventis 2010) or ethically charged issues (e.g., legalization of abortion; Inbar, Pizarro, and Bloom 2009). Conservatives prefer moral and legal wrongdoings to be punished more frequently and more severely than do liberals (Grasmick and McGill 1994; King and Maruna 2009). For instance, conservative U.S. voters punish (i.e., are less willing to reelect) their representatives in Congress when the latter support (on the basis of partisan loyalty) laws that go against their voters’ ideological positions on divisive issues (Carson et al. 2010).

In addition to being more punitive, conservatives are also less trusting. People with extremely conservative political views (e.g., ultraindividualists, procapitalists, state minimalists) exhibit low social and political trust (Williams 2020). Sunstein, Reisch, and Kaiser (2019) find that Americans who score higher on political conservatism are less likely to support nudges (i.e., implicit suggestions aimed at promoting welfare-maximizing behaviors by groups or individuals), even when they take the form of politically informed initiatives. Evans and Feng (2013) find that conservative Protestantism decreases willingness to follow scientists’ recommendations on issues such as climate change. Recently, these findings have been replicated in the context of the COVID-19 pandemic. Plohl and Musil (2021) find that conservatives place less trust in the global scientific community, ultimately leading to lower compliance with scientific guidelines aimed at preventing the spread of COVID-19.

Finally, conservatives exhibit stronger national attachment (Bealey 1999) and ethnocentric tendencies (Cunningham, Nezlek, and Banaji 2004). Conservative ideology has recently reconnected with populist demands for national sovereignty as well as contempt for supranational policies and deterritorialization of political power (Heinisch, Werner, and Habersack 2020). These facets of conservatism have been driving forces of increased skepticism toward deepening European integration and significant political developments such as Brexit (Hayton 2018). The conservatism-driven delegitimization of supranational governance should thus leave little space for international institutions to reconstruct trust bonds with conservative constituents during crises. Judged by conservatives as unable to prevent or manage nonlocal crises outside national institutions’ sphere of responsibility or range of control, international institutions constitute easy targets of blame and distrust.

H₃: Conservative political ideology (a) increases international institutional blame and decreases national institutional blame and (b) decreases international institutional trust and increases national institutional trust.

The “Citizens as Customers of Political Institutions” Paradigm and Well-Being

Over the past decades, public institutions have increasingly used advanced marketing techniques to address citizens’ needs (Walsh 1991). By closely monitoring public attitudes, institutions have been able to establish closer relationships with citizens to charge appropriately for the services they provide, address public needs in a superior way, and foster citizens’ satisfaction (Aberbach and Christensen 2005). For instance, governments use consumer relationship management tools to engage with the public, obtain insights, and encourage citizens to get involved in the coproduction of the public services they consume (e.g., services to disabled citizens), ensuring that citizens’ opinions and suggestions are heard (King 2018).

Essentially, the relationship between citizens and political institutions resembles a marketplace relationship between customers and firms. Similar to how consumers engage in marketplace exchanges with for-profit companies (whereby consumers offer financial resources for the acquisition of products or services), citizens engage in exchange relationships with nonprofit, political institutions responsible for their well-being (e.g., national governments, transnational governance bodies). The currency of these latter exchanges, though, is not financial; instead, it is relational and takes the form of institutional trust. Drawing parallels with conventional marketplace exchanges, we argue that citizens are on the receiving end of policies, public services, and institutional advice in exchange for the trust they place in the institutions providing these benefits. Much of the public policy literature (as well as emerging fields in psychology and behavioral economics such as nudging; Benartzi et al. 2017) relies on this paradigm. This transactional approach elevates institutional trust as a critical antecedent of the success of public policy initiatives and well-being guidance.

Institutional trust builds adherence to institutional guidelines. Trust is a prerequisite for advice acceptance from others, especially when those others are deemed experts in the matter for which advice is asked/offered (Sniezek and Van Swol 2001). The importance of trust when accepting expert advice is higher in conditions of social uncertainty, where information asymmetry often lurks (Kollock 1994). Global crises qualify as socially uncertain conditions, not least because of people’s inability to fully account for the causes and outcomes of events outside their “controlled” territory. Because crisis-regulating institutions are likely to hold such information, global crises relegate citizens to the role of advice seekers lacking perfect information and institutions to the role of expert advice givers, ultimately making trust of the former in the latter the determinant of advice acceptance and effectiveness.

The importance of institutional trust for the acceptance of public guidance has been shown across several contexts. Patients who trust their physicians are more likely to adhere to treatment plans agreed with them (Pearson and Raeke 2000). Trust in the police increases cooperativeness with police officers and compliance with the law (Jackson and Bradford 2010). Institutional trust increases recycling and waste management behaviors (Hansmann et al. 2006). Trust in environmental interest groups increases belief in climate change and adoption of environmentally friendly agricultural policies (Arbuckle, Morton, and Hobbs 2015). In the context of pandemics, trust in the national government boosted intention to follow vaccination recommendations after the H1N1 pandemic (Gilles et al. 2011). Similar effects apply for international institutions: the likelihood of voting for Brexit was higher among U.K. voters who distrusted the EU and opposed the advancement of European integration (Dustmann et al. 2017). In the health domain, Plohl and Musil (2021) find that individuals who trust the global scientific community respect the COVID-19 prevention guidelines more, in line with previous research on vaccination and trust in science (e.g., Keelan et al. 2010) showing that those who express more trust in science were also more willing to be vaccinated.

Beyond willingness to follow institutional advice, trust in national and international institutions benefit consumers’ subjective well-being. Using data from 15 European countries, Hudson (2006) finds that trust in national (i.e., local government and national law) and international institutions (e.g., the European Central bank, United Nations) relates positively with happiness measures. Analysis of data from the World Value Survey across 50 countries shows that confidence in national institutions such as the police, the judicial system, and the national government are strong predictors of citizens’ health, happiness, and life satisfaction (Elgar et al. 2011). The positive effects of institutional trust on happiness and well-being have been replicated in multiple countries, including Russia (Mironova 2015), Canada (Leung et al. 2011), and several East Asian nations (Yamaoka 2008). Although cross-country discrepancies in the strength of institutional trust’s influence on well-being exist—especially in transition economies and countries lacking well-established institutions (Jovanović 2016)—institutional trust (both national and international) is an important driver of well-being.

H₄: International institutional trust increases (a) adherence to international institutional guidelines and (b) consumer well-being.

OPEN IN VIEWER

Figure 1.

Institutional trust and adherence to well-being guidance as parallel transaction flows.

OPEN IN VIEWER

Figure 2.

Conceptual framework.

Empirical Context: The COVID-19 Pandemic

We use the COVID-19 pandemic as the empirical context to test our conceptual framework for three reasons. First, the pandemic has had severe negative consequences for international health care systems, the global economy, and consumer well-being. Antonio Guterres, the United Nations’ Secretary-General, has characterized the pandemic as the biggest challenge the world has faced since the Second World War. From its outbreak in China in December 2019 to the time of this writing, COVID-19 has caused approximately 150 million infections and claimed more than 3 million lives across 223 countries (WHO 2021). Second, despite its global impact, COVID-19 has affected countries around the world with different intensities. Infection numbers, hospitalization needs, and death rates vary significantly across countries (Chaudhry et al. 2020), suggesting high levels of “desirable” impact variance across countries. Third, the pandemic mobilized both international (e.g., the WHO, the global scientific community, pharmaceutical companies) and national institutions (e.g., national governments, health care systems), thus raising issues of coordinated governance between national and supranational regulatory actors and posing questions of institutional competence, legitimacy, and trust.

Overview of Empirical Studies

We conduct two complementary studies to test our predictions. Study 1 uses survey data collected in two countries during the pandemic’s first wave. In Study 1, we achieved variance in the independent variable by focusing on two countries hit by the pandemic with different intensity (the United States: high impact; Greece: low impact). We conducted Study 2 during the pandemic’s second wave, focusing on one country (Italy) to rule out cross-country confounds (e.g., differences in culture, economy, institutional environment). We created variance in perceived local impact through experimental manipulations that offer internal validity and causal (rather than just correlational) evidence in favor of our conceptual model.

Time of Data Collection and Country Selection

We collected survey data from citizens of two countries at the end of the pandemic’s first wave. Data were collected on September 17, 2020, in both countries. We opted for this date because the end of summer was considered the end of the pandemic’s first wave, when global rates of infection reached a short-term plateau for the first time after increasing exponentially since the pandemic’s outbreak (WHO 2021). We deemed avoiding data collection during the pandemic’s peak appropriate to ensure that (1) well-being scores are not clustered in the low end as a result of floor effects during lockdown periods and (2) sufficient time had passed since the pandemic’s outbreak for respondents to develop informed beliefs about the effectiveness of institutional policies designed to address the crisis.

We collected data in a high-impact country (the United States) and a low-impact country (Greece). We chose these two countries for three reasons. First, at the time of data collection, the United States was the country most severely hit by the pandemic in terms of infection rates and COVID-19-related fatalities. At the same time, Greece was among the few countries that prevented an uncontrolled local spread of the virus. Epidemiological metrics at the time of data collection indicate that (1) cumulative infections per million people (i.e., a measure allowing direct comparisons between countries with different populations) in the United States were almost 15 times higher than in Greece, (2) U.S. deaths attributed to COVID-19 stood at 600 per million people and the respective Greek figure barely reached 30, (3) U.S. daily infection and death rates were increasing at a much faster rate than in Greece, and (4) U.S. case fatality ratios (i.e., the proportion of cases resulting in fatal outcome) were slightly worse than Greece’s (Table 2). Second, despite differences in population, the United States and Greece (1) are both among the world’s 50 most developed countries in terms of gross domestic product (1st and 52nd, respectively) and Human Development Index (15th and 32nd, respectively) (World Bank 2020), (2) have similar political systems and engagement in supranational institutions (e.g., United Nations, North Atlantic Treaty Organization), and (3) have comparable cultural profiles. Such similarities minimize country-dependent discrepancies in institutional trust that countries with transitional economies and weaker institutional structures could cause. Finally, public perceptions about the handling of the pandemic in the United States and Greece differed substantially. Whereas the United States’ management of the pandemic has been described as a “tragedy” (The New York Times 2020), Greece, although considered an underdog in beating COVID-19, has been praised for its response to the pandemic’s first wave as a “great anomaly” (Giugliano 2020).

OPEN IN VIEWER

Table 2.

COVID-19 Epidemiological Metrics in the United States and Greece on Data Collection Date.

On September 17, 2020	United States (High Impact)	Greece (Low Impact)
Total cases	6,630,891	14,041
Total cases per million people	20,033	1,347
Total deaths	196,802	316
Total deaths per million people	595	30
Daily new cases	24,598	311
Daily cases per million people	74	30
Daily new deaths	865	3
Daily new deaths per million people	2.61	.29
Case fatality ratio	3.0%	2.3%

Notes: Data sourced from www.ourworldindata.org. Total figures refer to cumulative numbers from the start of the pandemic until the date of data collection. Daily figures refer to numbers on the date of data collection. Except for “Daily new deaths per million people” and “Case fatality ratio,” figures are rounded up to the nearest integer.

Participants and Method

Procedure and measures

Participants agreed to participate in a study exploring their perceptions about the impact of COVID-19 in their country. First, participants were asked to consider the overall impact of COVID-19 in their country and express the extent to which they believed their country was affected by COVID-19 compared with other similar countries: “I believe my country has suffered MORE than other similar countries because of COVID-19” (1 = “strongly disagree,” and 7 = “strongly agree”). We measured adherence to institutional guidelines with two equivalent three-item scales, one focusing on international guidelines and one focusing on national guidelines. We measured institutional trust using four items: “Most (inter)national institutions are basically honest, trustworthy, good and kind, and trustful of others” (Yamagishi and Yamagishi 1994). Participants answered these items once for international institutions and once for national institutions. Next, participants rated the extent to which they blamed international institutions (e.g., the WHO, European Union, United Nations) and national institutions (e.g., the national government, national health system, health ministry) for the way the COVID-19 pandemic evolved in their country. We used an adjusted version of Spears, Parker, and McDonald’s (2004) scale to measure globalization attitudes, enriched with some items intended to measure beliefs about globalization’s social benefits. We measured political orientation with the semantic differential question “Where would you place yourself on the following scale?” 1 = “very liberal,” and 7 = “very conservative” (Janoff-Bulman, Sheikh, and Baldacci 2008). Finally, we measured subjective well-being with five items adapted from Diener et al. (1985). All items were answered on seven-point scales. For the Greek sample, scales were translated and back-translated from English to Greek by a bilingual translator (Behling and Law 2000). Finally, participants provided demographic information, were debriefed, and thanked for their participation. Measurement scales and their psychometric properties are shown in Table 3.

OPEN IN VIEWER

Table 3.

Construct Measurement and Psychometric Properties (Study 1).

Construct	Psychometric Properties
Perceived Local Impact of Global Crisis (1 = “strongly disagree,” and 7 = “strongly agree”) I believe my country has suffered more than other similar countries because of COVID-19.	N.A. (single-item measure)
Institutional Blame (1 = “strongly disagree,” and 7 = “strongly agree”) I blame the NATIONAL/INTERNATIONAL institutions for how the COVID-19 situation evolved in my country.	N.A. (single-item measures)
International/National Institutional Trust (1 = “strongly disagree,” and 7 = “strongly agree”)	International institutions: α = .936; CR = .936; AVE = .785 National institutions: α = .946; CR = .946; AVE = .814
Most international/national institutions are basically honest.	λinternational = .895*** λnational = .921***
Most international/national institutions are trustworthy.	λinternational = .916*** λnational = .934***
Most international/national institutions are basically good and kind.	λinternational = .870*** λnational = .889***
Most international/national institutions are trustful of others.	λinternational = .862*** λnational = .863***
Adherence to International/National Institutional Guidelines (1 = “strongly disagree,” and 7 = “strongly agree”)	International institutions: α = .867; CR = .876; AVE = .704 National institutions: α = .868; CR = .881; AVE = .713
I will follow the guidelines of international/national institutions.	λinternational = .863*** λnational = .826***
I believe that the communications of international/national institutions are effective in preventing COVID-19.	λinternational = .740*** λnational = .772***
I would follow more comfortably the guidelines of the international/national institutions to stop the spread of the COVID-19.	λinternational = .905*** λnational = .927***
Subjective Well-Being (1 = “strongly disagree,” and 7 = “strongly agree”)	α = .900; CR = .906; AVE = .662
In most ways my life is close to ideal.	λ = .868***
The conditions of my life are excellent.	λ = .841***
I am satisfied with my life.	λ = .915***
So far, I have gotten the important things I want in life.	λ = .770***
If I could live my life over, I would change almost nothing.	λ = .649***
Political Orientation Where you would place yourself on the following scale: (1 = “very liberal,” and 7 = “very conservative”)	N.A. (single-item measure)
Globalization Attitude (1 = “strongly disagree,” and 7 = “strongly agree”)	α = .878; CR = .882; AVE = .557
Globalization encourages a maximum of personal freedom and choice.	λ = .729***
Globalization leads to quality and technical advances.	λ = .796***
Globalization provides consumers the goods and services they want.	λ = .724***
Globalization is more a force for good than bad.	λ = .812***
Globalization is good for the economy.	λ = .723***
Globalization is good for national cultures.	λ = .685**

Notes: N.A. = not applicable; CR = composite reliability; AVE = average variance extracted.

Analysis and Results

Invariance testing

We tested for configural and metric invariance to ensure that our measures’ factor structure is equivalent between the high- and low-impact country samples (Steenkamp and Baumgartner 1998). Regarding configural invariance, we conducted two CFAs, one for each country sample. The results indicate that the proposed factor structure fits the data collected in each country well, as evidenced by good model fit statistics, confirming configural invariance. Regarding metric invariance, we estimated both unconstrained (factor loadings set free to be estimated in each country sample) and constrained (factor loadings set equal across country samples) measurement models and tested for model fit changes. The χ² test results suggest that the constrained model’s fit is not significantly different from the unconstrained model’s fit, confirming metric invariance (Table 4).

OPEN IN VIEWER

Table 4.

Invariance Testing.

Configural Invariance	χ2	d.f.	RMSEA	CFI	SRMR	λs
United States (N = 198)	727.18	336	.077	.920	.048	all ps < .001
Greece (N = 196)	681.40	336	.073	.900	.054	all ps < .001
Total sample (N = 394)	1,047.48	336	.073	.914	.048	all ps < .001
Metric invariance	χ 2	d.f.	RMSEA	CFI	SRMR	λs
Unconstrained model	1,408.58	672	.073	.914	.048	λUSA, λGR freely estimated
Constrained model	1,438.60	691	.053	.910	.054	λUSA, λGR set equal
χ2 difference test	Δ(χ2) = 30.02, Δ(d.f.) = 19, p = .052

Notes: RMSEA = root mean square error of approximation; CFI = comparative fit index; SRMR = standardized root mean square residual.

Structural equations model estimation

We estimated our model through covariance-based structural equation modeling in AMOS using maximum likelihood estimation. Having established satisfactory measurement properties for all constructs (i.e., unidimensionality, reliability, and validity), we parceled the items of each construct and used the corresponding composites as single-item indicators of their latent variables (Bandalos 2002). To account for measurement error, we specified the error variances of composite indicators using the following formula: σ²_error = (1 – α) × σ²_composite, where α is the Cronbach’s alpha of the construct items and σ²_composite is the variance of the composite indicator (Anderson and Gerbing 1988).

The model fits very well to the data (χ² = 34.23, d.f. = 17, p = .008, RMSEA = .051, CFI = .987, SRMR = .025). Beyond satisfactory global fit, the model demonstrates very satisfactory local fit indices with reference to the hypothesized model paths (standardized parameters reported in text; for unstandardized estimates, see Table 5). The relative local impact of a global crisis has positive effects on both international institutional blame (β = .182, p < .01) and national institutional blame (β = .271, p < .001). Blame attributions exert negative effects on trust toward the corresponding institutions (β_{international} = −.333, p < .001; β_national = −.342, p < .001). Trust in national and international institutions increases willingness to follow well-being advice issued by national and international institutions, respectively (β_{international} = .490, p < .001; β_national = .469, p < .001). Regarding subjective well-being, we find that it increases as a function of national institutional trust (β = .276, p < .001) but not as a function of international institutional trust (β = .029, p = .732).

OPEN IN VIEWER

Table 5.

Structural Parameter Estimates (Studies 1 and 2).

	Endogenous Variables
	International Institutional Blame	National Institutional Blame	International Institutional Trust	National Institutional Trust	Adherence to International Guidelines	Adherence to National Guidelines	Subjective Well-Being	Well-Being Crisis Support
Independent Variables
H1: Local impact of global crisis	.186 (.061) ** .091 (.046) *	.286 (.061) *** .222 (.050) ***
H2: Globalization attitude	–.274 (.096) ** –.117 (.068) †	.013 (.097) .028 (.074)	.506 (.061) *** .322 (.056) ***	.400 (.071) *** .273 (.062) ***	.157 (.062) * .065 (.054)	.186 (.065) * .123 (.057) *	.186 (.081) * .123 (.072)†	N.A. –.042 (.044)
H3: Political ideology (conservative)	.182 (.060) ** .219 (.060) ***	–.199 (.061) ** –.022 (.065)	–.011 (.038) –.052 (.049)	.100 (.045) * .069 (.054)	–.134 (.035) *** –.101 (.045) *	–.114 (.040) ** –.172 (.049) ***	.124 (.046) ** .078 (.062)	N.A. –.027 (.038)
Serial Mediators
H1a: International institutional blame			–.240 (.024) *** –.398 (.036) ***
H1b: National institutional blame				–.259 (.027) *** –.335 (.038) ***
H4: International institutional trust					.432 (.042) *** .376 (.041) ***		.029 (.086) –.059 (.112)	N.A. –.048 (.069)
H5: National institutional trust						.406 (.039) *** .340 (.040) ***	.255 (071) *** .209 (.101) *	N.A. .230 (.062) ***
Controls
Age	.007 (.009) .020 (.009) *	–.011 (.009) .005 (.009)	.004 (.006) .012 (.007)†	.010 (.007) .010 (.008)	–.004 (.005) .002 (.006)	.008 (.006) –.001 (.007)	.001 (.007) .014 (.008)†	N.A. –.002 (.005)
Gender (1 = male, 2 = female)	–.075 (.166) .031 (.122)	.137 (.168) –.038 (.132)	–.229 (.105) * –.127 (.098)	–.241 (.123) * –.094 (.110)	.168 (.097)† .002 (.006)	–.024 (.110) –.001 (.007)	–.015 (.123) .014 (.008)†	N.A. –.002 (.005)
Education	N.A. –.019 (.053)	N.A. –.066 (.057)	N.A. .008 (.043)	N.A. .005 (.048)	N.A. .006 (.039)	N.A. .065 (.043)	N.A. .018 (.051)	N.A. –.042 (.031)
Socioeconomic status	N.A. .013 (.044)	N.A. –.048 (.048)	N.A. .129 (.036) ***	N.A. .046 (.040)	N.A. –.014 (.034)	N.A. .027 (.036)	N.A. .335 (.044) ***	N.A. .112 (.027) ***
COVID-19 personal impact index	N.A. –.010 (.086)	N.A. –.032 (.094)	N.A. –.006 (.070)	N.A. .002 (.079)	N.A. –.026 (.065)	N.A. .002 (.071)	N.A. .210 (.084) *	N.A. .108 (.051) *
Country dummy (1 = United States, 0 = Greece)	–.625 (.239) ** N.A.	.342 (.241) N.A.	.484 (.129) *** N.A.	.479 (.153) ** N.A.	–.380 (.121) ** N.A.	–.330 (.135) * N.A.	.004 (.155) N.A.	N.A. N.A.
R2	8.2% 8.5%	12.3% 6.5%	38.4% 34.6%	24.4% 19.6%	39.7% 24.6%	29.3% 22.5%	15.9% 22.9%	N.A. 17.3%
Model fit	Study 1: χ2 = 34.23, d.f. = 17, p = .008, RMSEA = .051, CFI = .987, SRMR = .025 Study 2: χ2 = 34.36, d.f. = 23, p = .060, RMSEA = .035, CFI = .991, SRMR = .023

^†

p < .10.

*p < .05.

**p < .01.

***p < .001.

Notes: Reported significances based on two tailed tests; all hypothesized effects significant (one-tailed) at α = 5%. Column entries refer to unstandardized parameters. Top (bottom) cell entries refer to Study 1 (Study 2) results. Standard errors are shown in parentheses. Numbers in boldface refer to hypothesized parameters. N.A. = not applicable; RMSEA = root mean square error of approximation; CFI = comparative fit index; SRMR = standardized root mean square residual.

Regarding globalization attitudes (H₂), we find a negative effect on international institutional blame (β = −.156, p < .01). Beyond the positive indirect effect on international institutional trust through decreasing international institutional blame, globalization attitudes have a direct positive effect on international institutional trust (β = .400, p < .001), suggesting partial mediation. Although we do not find a significant effect of globalization attitudes on national institutional blame (β = .007, p = .893), we observe a positive direct effect of globalization attitudes on national institutional trust (β = .291, p < .001), in support of H_2b. We also find positive direct effects of globalization attitudes on adherence to international guidelines (β = .141, p < .05), adherence to national guidelines (β = .157, p < .010), and subjective well-being (β = .146, p < .05). Turning to political ideology (H₃), conservative ideology has a positive effect on international institutional blame (β = .159, p < .01) and a negative effect on national institutional blame (β = −.168, p < .01), leading to positive and negative indirect effects on international and national institutional trust, respectively. Beyond indirect effects, conservative ideology has a positive direct effect on national institutional trust (β = .112, p < .05), suggesting partial mediation, but no direct effect on international institutional trust (β = −.013, p = .775), suggesting full mediation of its negative effects through blame. We also find negative direct effects of conservative ideology on adherence to international (β = −.185, p < .001) and national (β = −.147, p < .01) guidelines as well as a positive effect on subjective well-being (β = .150, p < .01). For indirect, direct, and total effect estimates of globalization attitudes and political ideology, we also used bootstrapping to obtain bias corrected 95% confidence intervals using 5,000 resamples (Table 6). The bootstrap estimates corroborate the normal theory-based estimates.

OPEN IN VIEWER

Table 6.

Indirect, Direct, and Total Effects of Globalization Attitude and Political Ideology (Studies 1 and 2).

Hypothesis	Effect	Mean Estimate	Standard Error	95% Lower Confidence Interval	95% Upper Confidence Interval	Hypothesis Result (Study 1)	Hypothesis Result (Study 2)
H1a	Local Impact → International Blame → International Trust (total indirect)	–.060/–.043	.022/.026	–.104/–.096	–.019/.006	Supported	Supported (one-tailed test)
H1b	Local Impact → National Blame → National Trust (total indirect)	–.093/–.086	.024/.025	–.143/–.137	–.051/–.041	Supported	Supported
H2a	Globalization Attitude → International Blame → International Trust (indirect)	.052/.042	.020/.030	.013/–.014	.093/.101	Supported (partial mediation)	Supported (direct effect)
	Globalization Attitude → International Trust (direct)	.400/.288	.051/.069	.295/.178	.495/.392
	Globalization Attitude → International Trust (total)	.452/.330	.055/.059	.333/.206	.549/.442
H2b	Globalization Attitude → National Blame → National Trust (indirect)	–.002/–.008	.020/.025	–.043/–.058	.036/.042	Supported (direct only)	Supported (direct only)
	Globalization Attitude → National Trust (direct)	.291/.236	.058/.068	.173/.121	.401/.349
	Globalization Attitude → National Trust (total)	.289/.228	.060/.063	.165/.104	.403/.349
H3a	Political Orientation → International Blame → International Trust (indirect)	–.053/–.082	.019/.028	–.092/–.138	–.018/–.029	Supported (indirect only)	Supported (full mediation)
	Political Orientation → International Trust (direct)	–.013/–.049	.051/.061	–.113/–.147	.086/.051
	Political Orientation → International Trust (total)	–.066/–.132	.052/.057	–.170/–.241	.032/–.021
H3b	Political Orientation → National Blame → National Trust (indirect)	.058/.007	.020/.022	.022/–.035	.101/.052	Supported (partial mediation)	Not supported
	Political Orientation → National Trust (direct)	.112/.063	.054/.059	.006/–.032	.218/.165
	Political Orientation → National Trust (total)	.170/.070	.056/.054	.053/–.031	.274/.180

Notes: Column entries refer to bootstrapped-obtained, bias corrected estimates (5,000 samples).

Finally, we observe effects of the country control dummy (1 = United States, 0 = Greece), which capture country idiosyncratic responses to the global crisis on top of any effects attributed to differences in perceived local impact (which are captured by the independent variable instead). Specifically, compared with Greek respondents, U.S. respondents (ceteris paribus) (1) blame international institutions less for the crisis (β = −.180, p < .01), (2) exhibit higher levels of national and international institutional trust (β_{international} = .185, p < .001; β_national = .169, p < .01), but (3) are less likely to follow national or international institutional guidelines (β_{international} = −.165, p < .01; β_national = −.134, p < .05). The results provide support to our conceptual model. Only H_4b is not supported, which indicates that international institutional trust is not a significant predictor of subjective well-being.

A typology of institutional responses during global crises

The results suggest that a substantial amount of variance in institutional blame, trust, and well-being is explained by consumers’ globalization attitudes and political ideology. Although, in theory, these constructs are expected to correlate—based on the premise that globalization is rooted in neoliberal ideology (Gill 2011)—our data suggest that such correlation (albeit significant) is rather small (r = −.224, p < .001), implying the potential to treat them as conceptually orthogonal. To further explore the combined effects of these constructs, we developed a typology of our respondents by identifying four distinct groups: conservative antiglobalists, conservative globalists, liberal antiglobalists, and liberal globalists. After allocating each respondent to one of these mutually exclusive categories according to self-reported scores of political ideology and globalization attitude (using the neutral scale points as classification thresholds), we developed a profile for each category by statistically comparing each group’s scores (using ANOVA tests) along the key model variables (Table 7).

OPEN IN VIEWER

Table 7.

Profiles of Institutional Responses Based on Political Ideology and Globalization Attitude.

	Conservative Antiglobalists	Conservative Globalists	Liberal Antiglobalists	LiberalGlobalists	Grand Mean	F	p
International institutional blame	5.09 (1.79)a	3.70 (1.75)b	3.62 (2.08)b	3.46 (1.72)b	3.68 (1.81)	8.59	<.001
	4.58 (1.83)	3.85 (1.94)	4.09 (2.15)	3.60 (1.68)	3.75 (1.79)	1.50	.216
	5.35 (1.75)a	3.65 (1.69)b	2.64 (1.57)b	3.27 (1.75)b	3.61 (1.84)	10.23	<.001
	3.76 (1.36)a	3.52 (1.42)a	3.10 (1.47)b	2.90 (1.23)b	3.11 (1.35)	6.74	<.001
National institutional blame	4.14 (1.88)	4.06 (1.82)	3.91 (2.00)	4.59 (1.86)	4.36 (1.88)	2.81	.039
	4.17 (1.70)	3.70 (1.84)	3.39 (1.88)	4.17 (1.75)	4.02 (1.78)	1.60	.191
	4.13 (2.00)b	4.20 (1.80)b	5.00 (1.90)a	5.19 (1.86)a	4.70 (1.91)	4.75	.003
	3.78 (1.52)	3.86 (1.52)	3.60 (1.42)	3.69 (1.43)	3.72 (1.45)	.45	.716
International institutional trust	2.96 (1.23)c	4.69 (1.22)a	3.79 (1.47)b	4.59 (1.22)a	4.40 (1.34)	22.05	<.001
	2.81 (1.07)b	4.36 (1.33)a	3.64 (1.42)b	4.21 (1.16)a	4.08 (1.26)	6.37	<.001
	3.03 (1.32)c	4.82 (1.16)a	4.09 (1.61)b	5.13 (1.08)a	4.72 (1.51)	20.95	<.001
	3.74 (1.18)b	3.93 (1.15)b	3.80 (1.42)b	4.62 (1.14)a	4.30 (1.24)	13.79	<.001
National institutional trust	3.16 (1.38)c	4.53 (1.28)a	3.49 (1.52)b	3.90 (1.44)b	3.96 (1.45)	10.72	<.001
	2.90 (1.32)b	4.22 (1.21)a	3.48 (1.42)a	3.68 (1.33)a	3.68 (1.35)	3.09	.028
	3.29 (1.42)b	4.65 (1.29)a	3.50 (1.79)b	4.22 (1.53)a	4.22 (1.51)	6.11	.001
	3.76 (1.13)a	3.65 (1.14)a	3.25 (1.37)b	3.94 (1.28)a	3.77 (1.27)	4.93	.002
Adherence to international institutional guidelines	4.19 (1.57)b	5.35 (1.09)a	5.32 (1.42)a	5.66 (1.03)a	5.42 (1.21)	17.01	<.001
	4.78 (1.73)b	5.49 (.81)a	5.35 (1.49)a	5.61 (1.00)a	5.51 (1.11)	2.32	.077
	3.88 (1.42)b	5.30 (1.18)a	5.27 (1.33)a	5.72 (1.08)a	5.33 (1.29)	15.10	<.001
	5.22 (1.13)b	5.44 (1.10)a	5.66 (1.22)a	5.75 (1.03)a	5.64 (1.09)	3.08	.027
Adherence to national institutional guidelines	4.56 (1.74)b	5.32 (1.12)a	5.15 (1.36)a	5.38 (1.26)a	5.27 (1.30)	4.27	.006
	4.58 (1.81)	5.40 (.82)	5.04 (1.40)	5.61 (1.00)	5.51 (1.11)	1.99	.117
	4.55 (1.74)b	5.29 (1.22)a	5.36 (1.29)a	5.34 (1.27)a	5.23 (1.33)	2.34	.075
	4.94 (1.31)b	5.43 (1.14)a	5.67 (1.24)a	5.78 (1.09)a	5.64 (1.16)	5.57	.001
Subjective well-being	4.04 (1.46)b	4.74 (1.36)a	3.57 (1.31)b	4.10 (1.33)b	4.21 (1.38)	8.41	<.001
	3.85 (1.19)	4.56 (1.44)	3.71 (1.27)	4.02 (1.23)	4.05 (1.27)	2.16	.094
	4.14 (1.60)a	4.81 (1.32)a	3.27 (1.41)b	4.20 (1.46)b	4.40 (1.47)	5.06	.002
	3.96 (1.14)	3.60 (1.41)	3.55 (1.43)	4.01 (1.42)	3.93 (1.41)	1.64	.179

Notes: Cell entries correspond to mean scores (standard deviations shown in parentheses). The first number in each cell corresponds to the pooled sample (Study 1), the second to the Greek sample (Study 1), the third to the U.S. sample (Study 1), and the fourth to the Italian sample (Study 2). Superscripts refer to statistically significant Bonferroni post hoc comparisons; “a” refers to the highest mean across cells, “b” refers to the second highest mean, and so on.

The results show interesting differences in crisis reactions by group. Conservative antiglobalists (8.9% of the sample; M_age = 37 years, 60% male) blame international institutions more for the mismanagement of the crisis, exhibit high levels of distrust in both national and international institutions, and report the lowest intention to follow their health guidelines. Conservative globalists (24.9% of the sample; M_age = 33 years, 65.3% male) assign equal levels of blame to both national and international institutions and report the highest level of institutional trust among all groups. They also report high levels of well-being and adherence to institutional guidelines. Liberal antiglobalists (8.6% of the sample; M_age = 32 years, 58.8% male) blame national and international institutions to similar extents but report low levels of trust, especially in national institutions. Despite not trusting international institutions strongly, they follow health advice issued by these institutions. They also report the lowest levels of well-being among groups. Finally, liberal globalists (57.6% of the sample; M_age = 29 years, 60.4% male) blame national institutions, report high levels of international institutional trust, and have the strongest adherence to institutional guidance.

We tested the typology’s stability by conducting the analyses in the high- and low-impact country samples separately. Although the results are largely robust across samples, we observe some noteworthy differences. First, U.S. respondents were more pessimistic about the future of the pandemic, as they reported significantly higher scores on a scale measuring their expectations about the persistence of the crisis in the long run (M_USA = 4.21, SD = 1.12 vs. M_GR = 3.66, SD = 1.24, t = 4.573, p < .001). Second, the U.S. sample reported higher variance in blame, trust, and well-being scores, suggesting intense polarization in U.S. citizens’ perceptions about who is at fault for the crisis and who should be trusted to guide them out of it. Finally, the U.S. sample was comprised of disproportionately higher percentages of conservative globalists and antiglobalists than the Greek sample, likely explaining the increased frequency of passive (e.g., rejection of health guidance, neglect of hygiene advice) and active (e.g., antilockdown protests, antimask rallies) anti-institutional behaviors in the United States (similar to other countries heavily impacted by the pandemic).

Time of Data Collection and Country Selection

We conducted an online experiment with a sample of Italian consumers. Data were collected during the period February 5–10, 2021. On the dates of data collection, Italy was experiencing the second wave of the pandemic, with an average of 12,000 new cases and 360 deaths per day. Collecting data during this period offers the opportunity to draw contrasts with the findings of Study 1, observe how the patterns of the hypothesized relationships had changed across the two waves of the pandemic, and assess the stability of our conceptual model over time.

Because we opted for an experimental manipulation of respondents’ perceived relative local impact in Study 2, we had to choose a country for which both relatively high and relatively low impact perceptions could be manipulated in a credible manner. Italy served this purpose as it was one of the countries that was heavily affected by the pandemic’s first wave (thus, perceptions of high local impact can be easily primed) but also one of the countries that did far better than other similar countries (e.g., the United Kingdom, France, Spain) during the second wave of the pandemic (thus, perceptions of low local impact can also be credibly construed).

Participants and Method

Analysis and Results

Measurement model assessment

A CFA indicates good fit of the model to the data (χ² = 1,194.76, d.f. = 524, RMSEA = .057, CFI = .926, SRMR = .045) and sound measurement for all multi-item scales. All scales have satisfactory psychometric properties, including Cronbach’s alphas (ranging from .801 to .906), CRs (ranging from .811 to .907), and AVEs (ranging from .467 to .749). In support of discriminant validity, all squared correlations are smaller than the minimum of the corresponding AVEs (Fornell and Larcker 1981) (see Table 8).

OPEN IN VIEWER

Table 8.

Construct Measurement and Psychometric Properties (Study 2).

Construct	Psychometric Properties
Perceived Local Impact of Global Crisis (1 = “strongly disagree,” and 7 = “strongly agree”) I believe my country has suffered more than other similar countries because of COVID-19.	N.A. (single-item measure)
Institutional Blame (1 = “strongly disagree,” and 7 = “strongly agree”) I blame the …… for how the COVID-19 situation evolved in my country.	N.A. (single-item measures)
International Institutions	α = .838; CR = .838; AVE = .643
World Health Organization	λ = .802***
European Union	λ = .810***
United Nations	λ = .776***
National Institutions	α = .834; CR = .847; AVE = .655
National government	λ = .875***
National health system	λ = .626***
Ministry of Health	λ = .898***
International/National Institutional Trust (1 = “strongly disagree,” and 7 = “strongly agree”)	International institutions: α = .906; CR = .907; AVE = .710 National institutions: α = .905; CR = .905; AVE = .705
Most international/national institutions are basically honest.	λinternational = .901*** λnational = .888***
Most international/national institutions are trustworthy.	λinternational = .911*** λnational = .891***
Most international/national institutions are basically good and kind.	λinternational = .774*** λnational = .797***
Most international/national institutions are trustful of others.	λinternational = .775*** λnational = .777***
Adherence to International/National Institutional Guidelines (1 = “strongly disagree," and 7 = “strongly agree”)	International institutions: α = .888; CR = .894; AVE = .737 National institutions: α = .888; CR = .899; AVE = .749
I will follow the guidelines of international/national institutions.	λinternational = .873*** λnational = .880***
I believe that the communications of international/national institutions are effective in preventing COVID-19.	λinternational = .783*** λnational = .769***
I would follow more comfortably the guidelines of the international/national institutions to stop the spread of the COVID-19.	λinternational = .915*** λnational = .939***
Subjective Well-Being (1 = “strongly disagree, and 7 = “strongly agree”)	α = .897; CR = .898; AVE = .689
In most ways my life is close to ideal.	λ = .895***
The conditions of my life are excellent.	λ = .811***
I am satisfied with my life.	λ = .858***
So far, I have gotten the important things I want in life.	λ = .770***
Consumer Well-Being Support To which extent did you receive support from national (e.g., national government, national health system, health ministry) and international (e.g., World Health Organization, European Union, United Nations) institutions during the COVID-19 pandemic? (1 = “far too little support, and 7 = “far too much support”)	α = .801; CR = .811; AVE = .467
Health care support	λ = .542***
Financial support	λ = .599***
Psychological support	λ = .764***
Social support	λ = .690***
Legal support	λ = .789***
Political Orientation Where you would place yourself on the following scale: (1 = “very liberal,” and 7 = “very conservative”)	N.A. (single-item measure)
Globalization Attitude (1 = “strongly disagree,” and 7 = “strongly agree”)	α = .875; CR = .882; AVE = .559
Globalization encourages a maximum of personal freedom and choice.	λ = .768***
Globalization leads to quality and technical advances.	λ = .801***
Globalization provides consumers the goods and services they want.	λ = .799***
Globalization is more a force for good than bad.	λ = .782***
Globalization is good for the economy.	λ = .750***
Globalization is good for national cultures.	λ = .555**

Theoretical Contribution

Our findings add to crisis management, institutional trust, and consumer well-being literatures. We discuss these contributions further in the subsections that follow.

Practical Implications

Our findings have implications for policy makers, political marketers, crisis communicators, and for-profit organizations actively engaged in global crisis management. We discuss each in turn in the following subsections.

Limitations and Future Research

Our findings are subject to limitations, which offer future research directions. First, the particularity of our empirical context (i.e., the COVID-19 pandemic) limits bold generalizations of our findings to other types of global crises with different qualitative characteristics. Although we expect the identified theoretical mechanism of institutional trust building, and the role of the global–local contrast in crisis impact, to hold in the context of other global crises, every crisis is unique in terms of both its causes (e.g., national vs. international origin, controllable vs. uncontrollable) and its national or international intervention potential (e.g., addressing a crisis’s global causes vs. handling its local symptoms). Similarly, differences in the severity of a global crisis in terms of its impact (e.g., human lives lost due to the pandemic vs. global terrorist acts) or the crisis’s general domain (e.g., financial disruptions vs. environmental disasters) could lead to different conclusions. Thus, replications and extensions of the current study in other crisis contexts would be useful.

Second, our studies focused on individual-level drivers of global crisis reactions. However, countries with different cultural or economic profiles likely react to global crises differently. For example, low-power-distance cultures may experience smaller declines in trust than high-power-distance cultures. Citizens of more economically developed countries may perceive their governments as more capable (and more trustworthy) in managing global crises locally. Countries with high percentages of vulnerable consumers (e.g., bottom of the pyramid markets) might also react to global crises differently. Future research should investigate country-level determinants of institutional (dis)trust following global crises.

Third, our study assesses trust and well-being at two distinct and incomparable points in time following a crisis outbreak. However, global crises are ongoing challenges that evolve over time, and institutional assessments of crisis management policies change continuously as crises unfold. Longitudinal studies would allow investigation of the trust dynamics developed during a developing crisis and the long-term effects of national and international interventions on institutional blame, trust, and citizens’ well-being.

Fourth, global crises offer a unique context to study the relative political legitimacy of national and international institutions, especially when their crisis guidelines are at odds or in outright conflict (e.g., countries following a “herd immunity” approach to the COVID-19 pandemic against WHO guidelines). Divergence in global crisis policies is not uncommon (e.g., anti-immigration policies of populist EU governments opposing EU immigration directives, the decision of the Trump administration to exit the United Nations’ Paris climate agreement). How trust differences in national governments versus supranational organizations affect intentions to follow national or international guidelines when they are in direct conflict is an area of significant public policy relevance that future research should explore in more depth.

Finally, future research should investigate how global crises influence consumption-related outcomes indirectly linked with well-being (decisions to invest in, e.g., education, property, insurance, pensions) as well as consumers’ dispositions toward globalization (e.g., global/local identities, consumer ethnocentrism, cosmopolitanism; Diamantopoulos et al. 2019) as a force with the potential to threaten and improve their well-being simultaneously.