Institutional and cultural determinants of speed of government responses during COVID-19 pandemic

The study investigates why governments responded at different speeds to the COVID-19 pandemic, treating the global outbreak as a natural experiment. It focuses on cultural (individualism–collectivism and power distance) and institutional (democracy and media freedom) predictors of response speed, defined as the marginal rate of change in the Oxford COVID-19 Government Response Stringency Index. The authors posit that individualistic cultures may resist swift restrictive measures due to higher valuation of personal freedom, while collectivistic cultures may accept restrictions more readily. They also explore whether trust in government moderates cultural effects on policy speed. The importance lies in understanding determinants of rapid policy action during novel crises where standard responses are lacking and where restrictive measures entail significant trade-offs in freedom, economic activity, and wellbeing.

Institutional factors: The role of democracy and media freedom in crisis response is theoretically ambiguous. Democracies may respond faster due to electoral accountability and media scrutiny, yet decision-making can be slower due to pluralistic deliberation and jurisdictional fragmentation. Prior studies on COVID-19 provide mixed evidence regarding the speed and effectiveness of democratic versus autocratic responses, and about the influence of free media on responsiveness and compliance. Cultural factors: Two Hofstede dimensions are emphasized—individualism vs. collectivism and power distance. Individualistic societies prize personal freedom and may resist stringent measures, while collectivistic societies exhibit higher willingness to sacrifice personal freedom for collective welfare. Power distance may facilitate acceptance of top-down stringent policies. Trust in government: Trust is expected to moderate cultural effects—high trust strengthens the alignment of government action with cultural orientations (e.g., quicker action in collectivistic, high-trust contexts; more hesitation in individualistic, high-trust contexts). Empirical exploration proceeds without definitive ex-ante hypotheses for democracy and media freedom due to theoretical ambiguity.

Design: Cross-country empirical analysis using the Oxford COVID-19 Government Response Tracker (OxCGRT) Stringency Index from 2020-01-01 to 2020-04-30. Dependent variable: Stringency Speed (SP), defined as the marginal rate of change of the stringency index, standardized by the country’s initial (2020-01-01) and maximum index within the period, then multiplied by 100 to improve scale. An alternative measure of proactivity (PRO) is the average stringency from the start of increase to its maximum within the same period; PRO correlates with SP at r = 0.69 (p < 0.01). Variables: Institutional—Democracy (Economist Intelligence Unit), Press Freedom (Reporters Without Borders). Cultural—Individualism (IDV) and Power Distance (PDI) from Hofstede. Moderator—Government trust (World Values Survey; percent reporting high confidence in government). Controls—ln(GDP per capita), ln(Population), Healthcare Access and Quality Index (HAQ), ln(days to first confirmed case since 2020-01-01), ln(days to first death), plus continent fixed effects. Data coverage: Up to 152 countries for main descriptive statistics; regression models have 71 observations due to data availability (e.g., trust). Estimation: OLS regressions with robust standard errors and continent fixed effects; additional robustness includes a dummy for prior severe epidemics (SARS, MERS, Ebola) which was not significant and did not alter main results. Further analyses examine policy-specific speed correlations (e.g., school closures, public transport) and downstream associations between SP (as of 2020-04-30) and COVID-19 deaths and cases per million (as of 2021-01-10), with subgroup splits by median individualism and government trust (four groups).

• Descriptives: Stringency Speed mean 0.3 (SD 0.3; N = 152), indicating substantial cross-country variation; highest values in East and Southeast Asia; Italy and Germany among Europe’s highest.
• Policy-specific correlations (Table 3): Democratic and individualistic countries are slower to close public transport (Democracy r = -0.287***; Individualism r = -0.103) and individualistic countries are slower to close schools (r = -0.228**). Individualism is strongly negatively correlated with speed of stay-at-home orders (r = -0.326***).
• Regression results (Table 4; N = 71; continent fixed effects; robust SEs):
  • Individualism predicts slower Stringency Speed across specifications (negative and statistically significant coefficients), indicating that higher individualism is associated with slower government response.
  • Democracy, press freedom, and power distance do not show significant predictive power for Stringency Speed in the authors’ interpretation, consistent with theoretical ambiguity for these factors in novel crises.
  • A moderating pattern by government trust is observed: the negative association of individualism with speed is stronger in higher-trust countries (illustrated in Fig. 2).
• Outcomes association (Table 5; as of 2021-01-10): Overall, faster Stringency Speed is associated with lower COVID-19 deaths and confirmed cases per million (e.g., death rate coefficient ≈ -2.43***; cases coefficient ≈ -2.80***). Subgroup patterns differ:
  • Collectivistic countries (low individualism), regardless of trust level: faster speed relates to fewer deaths and cases.
  • High-trust, high-individualism: faster speed relates to fewer deaths (significant), but not to fewer cases (possibly due to higher testing).
  • Low-trust, high-individualism: no significant relationship between speed and deaths or cases.
• Robustness: An alternative proactivity measure (PRO) corroborates main results; adding a prior-epidemic-experience dummy does not change findings.

Findings indicate that culture, particularly individualism versus collectivism, is a primary determinant of the speed of introducing stringent COVID-19 policies. Governments in more individualistic societies reacted more slowly, whereas those in collectivistic cultures reacted faster. Trust in government strengthens the alignment between cultural orientation and policy speed: in high-trust contexts, governments may cater more to citizens’ cultural preferences, increasing hesitation in individualistic societies and facilitating swift action in collectivistic ones. Institutional factors such as democracy and media freedom did not robustly predict speed, reflecting their theoretically ambivalent roles in novel crises that require swift, potentially contentious restrictions. The relationship between speed and public health outcomes varies by cultural-trust context: faster action is broadly associated with fewer deaths and cases, but effectiveness is muted where individualism is high and trust is low. Policy implications include the need for strategies that build trust and encourage collective orientations to enhance the effectiveness of rapid interventions, and the suggestion that even in individualistic high-trust societies, quicker action can reduce deaths, implying potential benefits of a more paternalistic stance during such crises.

The study shows that cultural factors—especially individualism versus collectivism—dominate institutional factors in explaining cross-country variation in the speed and proactivity of COVID-19 policy responses. Higher individualism is associated with slower responses, and this effect is stronger where government trust is high. Democracy, media freedom, and power distance do not robustly predict speed in this context. Faster responses are generally linked to better pandemic outcomes (lower deaths and, often, cases), though effectiveness depends on the interplay of culture and trust. Future research should test these relationships in other novel crises, investigate additional cultural dimensions (e.g., uncertainty avoidance, long-term orientation), and probe mechanisms—particularly compliance—by linking national culture to individual-level behavioral data.

• Causality: Cross-sectional design limits causal inference between cultural/institutional factors and response speed.
• Measurement: The OxCGRT stringency index, while comprehensive, is a coarse measure that may miss policy nuances; definitions of speed could vary.
• Sample constraints: Missing data (e.g., trust) reduce regression sample size; subgroup analyses have small Ns (<30 in some groups), weakening statistical validity.
• Dynamics: Government trust can change during crises, complicating interpretation of its moderating role measured pre-pandemic.
• Experience heterogeneity: Differences in prior epidemic experience may matter, though a simple dummy was not significant in robustness checks.