Why is trust key to managing crises?

Covid-19 has demonstrated how future upheavals – from pandemics to climate change – will require strong cooperation between all actors, public and private. This column argues that trust, whether between people or in government and scientists, is a critical factor in addressing the challenges of a crisis like Covid-19. While health characteristics explain one-quarter of the cross-country heterogeneity in a combined index of GDP growth and health outcomes during the crisis, trust in government alone accounts for two-thirds of the variance.

Covid-19 heralds the crises of the 21st century. From pandemics to climate change, future upheavals will require strong cooperation between all actors – public and private. This is the main message of our analysis of the Organisation for Economic Co-operation and Development (OECD)’s cross-country performance in the management of the Covid-19 crisis. Building on a set of analyses presented in Algan and Cohen (2021), we argue that trust of different kinds – be it interpersonal trust or trust in government and scientists – is a critical factor in addressing the challenges of a crisis like Covid-19. While health characteristics explain one-quarter of the cross-country heterogeneity of a combined index of GDP growth and health outcomes, trust in government alone accounts for two-thirds of the variance. Our paper contributes to the growing literature on the main features of crisis management across countries (Brodeur et al. 2021, Boone and Ladreit 2021, Maloney and Taskin 2020, Fan et al. 2020) by analysing the trust factor.

From one semester to the next: The evolution of the relationship between economic and health outcomes during the pandemic

How can we explain the cross-country heterogeneity in the management of the COVID-19 crisis, and its health and economic consequences?

At first sight, the compliance of citizens with restrictive measures has been central to the successful limitation of Covid-19 consequences, in terms of health and economic outcomes. This connection has been effectively documented, particularly by Barrios et al. (2021), who showed how civic capital induced higher levels of social distancing, as well as by Durante et al. (2021) and, in the French context, by Aminjonov and Bargain (2020).

While the economic consequences of past epidemics tended to be proportional to the number of deaths, the Covid-19 outbreak has gradually evolved to become a situation in which the two factors ceased to be correlated. In the first six months of the Covid-19 outbreak, we do observe a strong correlation between the magnitude of the pandemic and the evolution of GDP at the country level. The economic shrinkage was stronger in countries where the virus was widely spread and deadly. But this relationship tends to disappear during the following months and become rapidly statistically insignificant, until the health aspect of the crisis no longer explains the economic outcome (Figure 1).

Figure 1

Note: in the first half of 2020 in France, there were 457 deaths per million inhabitants and a recorded fall in GDP of around -13% compared to the same period in 2019. Lines indicate the correlation relationship for each period.

In France, we observe that the number of deaths per inhabitants is relatively high and stable over the three periods, at slightly more than 400 deaths per million inhabitants. The first half of 2020 is associated with one of the most important economic downturns (around –13% relative to the same period in 2019). During the second half of 2020 and the beginning of 2021, for the same number of deaths, the economic outcome is around –4 % relative to 2019, which suggests that we learned how to deal with the virus economically. The UK sees similar results, as the number of deaths in the first half of 2021 is slightly higher than the previous period, but the change in GDP is contained at around –7%. The US sits above the regression line, illustrating the choice to not impose a national lockdown.

The econometric analysis confirms the prominent role played by epidemic-related variables to explain GDP’s fall in the first semester (Figure 2). During the second semester, the correlation weakened while social variables (trust in others and in governments) became more salient, as confirmed by the first quarter of 2021. Comparing the relative explanatory power of the different set of variables, we find that health variables alone explain around 25% of the total variance, rising to 64% when trust factors are added.

Figure 2

Note: The coefficients from ordinary least squares regressions are shown. All variables are standardised (centred-reduced). The dot indicates the value of the coefficient and the bars indicate confidence intervals.

As for mortality, the econometric analysis shows that socio-demographic and health-related factors were determinants to explaining differences in the magnitude of the death toll (Figure 3). In particular, we analysed the age structure and density of the population as well as the number of hospital beds per 1,000 inhabitants. In the first semester, those factors are crucial to understanding death toll differentials; they become statistically insignificant in the second semester of 2020. On the other hand, while social variables (interpersonal trust and confidence in government) appear to have been weakly linked to mortality in the first semester, they become highly significant to explaining the death toll of the second semester. The explanatory power of socio-demographic variables in the first half of 2020 and of trust variables in the second half of 2020 are comparable, suggesting a major change in the dynamic of the crisis.

Figure 3

Note: The coefficients from ordinary least squares regressions are shown. All variables are standardised (centred-reduced). The dot indicates the value of the coefficient and the bars indicate confidence intervals.

We use those results as a basis to build a comparative index of outcomes by country, accounting for both the economic and the health dimensions of the crisis. We build a ‘sacrifice index’, adding and weighting the GDP change observed during separate periods and the death toll for each country during the period overall.¹ Then, we correlate the sacrifice index to better understand what was at stake during the crisis (Figure 4), using the degree of confidence in governments measured by the World Values Survey before the crisis began.² It appears that the correlation between the sacrifice index and the pre-crisis level of confidence in government is positive and strong. Considering government trust in our set of countries as a more general measure of the quality of the political environment – the reciprocal confidence between governments and citizens – this correlation confirms the central role of trust in the outcome of the pandemic. For example, low levels of trust imply a demand (from both governments and people) for more restrictive measures, hitting the economy harder.

Building on this statistical relationship, we observe that some countries do better and other do worse than their level of confidence in government would predict. France is precisely on the correlation line but the UK is below, reflecting the government’s erratic response to the first wave, as are Italy and Spain. At the opposite end of the graph, the Nordic countries, with a very high level of trust, performed better, as did the three main ‘zero-covid strategy’ countries (Australia, Korea, and New Zealand).

Figure 4

Trust in scientists is also strongly linked with the collective management of the pandemic. Following individuals in 12 countries, Algan et al. (2021) show that trust in scientists is strongly correlated to support of restrictive measures and to compliant behaviour. It is also linked with decisions around vaccinations. Trust in scientists evolved during the pandemic, and the drop in trust in scientists observed in France translated into much lower support for restrictive measures, as observed during different waves of the survey. One factor likely to explain the weakening of the trust in scientists is educational levels, particularly in science. Indeed, countries where the Programme for International Student Assessment (PISA) tests in sciences are the lowest saw more of a drop in trust in scientists during the pandemic, particularly Italy, the US, and – of a broader magnitude – France.

Figure 5

Trust is a central variable in the analysis of the Covid-19 crisis. Horizontal interpersonal trust, or vertical trust in government or scientists appear to be critical factors in understanding how societies reacted to the pandemic and which health and economic outcomes they managed to produce.

Algan, Y and D Cohen (2021) “The French in the Time of Covid-19 : an Economy and Society facing the Health Risk”, Notes du Conseil d’Analyse Economique 66: 1–12.

Algan, Y, D Cohen, E Davoine, M Foucault and S Stantcheva (2021), “Trust in scientists in times of pandemic: Panel evidence from 12 countries”, VoxEU.org, 15 December, and Proceedings of the National Academy of Sciences 118(40).

Bargain, O and U Aminjonov (2020), “Trust and compliance to public health policies in the time of COVID-19”, VoxEU.org, 23 October, and Journal of Public Economics 192: 104316.

Barrios, J M, E Benmelech, Y V Hochberg, P Sapienza and L Zingales (2021), “Civic capital and social distancing during the Covid-19 pandemic”, Journal of Public Economics 193: 104310.

Boone, L and C Ladreit (2021), “Fear of COVID and non-pharmaceutical interventions: An analysis of their economic impact among 29 advanced OECD countries”, VoxEU.org, 23 October.

Brodeur, A, D Gray, A Islam and S Bhuiyan (2021), “A literature review of the economics of COVID‐19”, Journal of Economic Surveys 35(4): 1007–1044.

Durante, R, L Guiso and G Gulino (2020), “Civic capital and social distancing: Evidence from Italians’ response to COVID-19”, VoxEU.org, 16 April, and Journal of Public Economics 194: 104342 (2021).

Fan, Y, A Yeşim Orhun and D Turjeman (2020), “Heterogeneous actions, beliefs, constraints and risk tolerance during the COVID-19 pandemic”, NBER Working Paper No. 27211.

Jones C J, T Philippon and V Venkateswaran (2020), “Optimal Mitigation Policies in a Pandemic: Social Distancing and Working From Home”, NBER Working Paper No. 26984.

Ladreit C (2021), “Government rule or fear of COVID? Which had more impact on economic activity?”, VoxEU.org, 06 April.

Maloney, W F and T Taskin (2020), “Voluntary vs mandated social distancing and economic activity during COVID-19”, VoxEU.org, 15 May, and World Bank Policy Research Working Paper (9242).

Péron M (2021), “Analyses d’une crise : éléments quantitatifs sur le choc Covid-19”, Focus du CAE No. 66, October.

1 We actually divide the mortality factor by 100, which amounts to taking the value of life to be equal to 100 times the income per head. The index is robust to alternative weighting method that would give an implicit value around 60 times the average income, as estimated by Jones et al. (2020).

2 The exact question is: “I am going to name a number of organizations. For each one, could you tell me how much confidence you have in them: is it a great deal of confidence, quite a lot of confidence, not very much confidence or none at all? : The government (in your nation’s capital)”.