The political consequences of the Covid pandemic: Lessons from cross-country polling data

Helios Herrera, Max Konradt, Guillermo Ordoñez, Christoph Trebesch 06 November 2020

The Covid-19 pandemic put governments across the world under pressure to react quickly and decisively. Their policy responses, however, varied substantially across countries. Some governments quickly imposed strict lockdown policies to keep case numbers in check (e.g. Australia or Argentina). Others opted for loose policies to reduce the economic damage of the pandemic (e.g. Brazil, Sweden, or the US). This trade-off between ‘health versus the economy’ has been widely debated in recent months (Alvarez et al. 2020, Eichenbaum et al. 2020, Farboodi et al. 2020). Yet, so far, there is still limited systematic evidence on how the public evaluates the different policy reactions for several countries and over several months.

This column, based our recent research (Herrera et al. 2020), studies the political consequences of (mis-)managing the Covid-19 pandemic. We ask the following questions: How does a government’s handling of the pandemic affect its political approval, and thus its re-election chances? Do governments get punished politically if they fail to respond strongly or promptly (or if they see infections and fatalities raise)? And what does the public care more about – good or bad news about infection case numbers, or news concerning the economy?

To address these questions, we build a novel international, high-frequency polling dataset, which consists of surveys on leaders’ approval and voting intentions on a weekly basis for 35 countries. Our dataset includes 20 advanced economies and 15 emerging market economies for which frequent, high-quality polling data was available (see Herrera et al. 2020b for a similar approach). We investigate how Covid-19 infection and fatality numbers affect approval rates over time, while controlling for government pandemic policies (using the weekly ‘Oxford stringency index’), as well as for economic activity (using weekly mobility and electricity data). The high-frequency panel structure of our dataset is key for capturing the dynamics of leaders’ approval, and represents an innovation as related studies typically rely on one-time election results, one-time survey data, or dynamics in a single country (Bol et al. 2020, Giommoni and Loumeau 2020). 

Our main finding, while seemingly intuitive, offers important political lessons. We find that governments are ‘punished’ in terms of political approval when infection numbers accelerate. This finding, however, only holds for governments that fail to impose stringent counter-measures. Moreover, we do not find that approval rates react to high-frequency measures of economic activity in this pandemic. Overall, this evidence suggests that loose pandemic policies are politically costly. Governments that placed more weight on health rather than short-term economic outcomes get rewarded in their approval. 

Large heterogeneity in approval changes during the Covid-19 crisis

As for the first months of the pandemic, Figure 1 graphs the percentage change in approval between February and July 2020 for the countries in our sample (ranked in descending order). On average, countries increased their approval by 16% (or 4.8 percentage points) during that period. The ranking is led by Australia, whose prime minister Scott Morrison saw his approval level increase by 90% (31 percentage points). Governments with the strongest fall in approval include Romania, Japan, and the US, where, as of July, President Trump had seen his approval rating drop by approximately 17% (eight percentage points) compared to pre-pandemic levels.

READ ALSO  Vanguard Real Estate ETF: Cost-Efficient ETF With A Reasonable Income Stream (NYSEARCA:VNQ)

Figure 1 Government approval during the Covid-19 crisis, change from February to July 2020 (%)

Note: This figure shows the percentage change in government approval from February to July 2020 for the 35 countries in our sample. The approval data build on a newly collected dataset combining political polls on leader approval and polls on voting intentions for the coalition government parties.

Governments which fail to contain the virus are punished politically

Our main finding is a negative relationship between Covid-19 infection numbers and political approval across countries and time. Controlling for deaths and a proxy for economic activity (and including time and country fixed effects), we show that a one standard deviation increase in weekly case growth is associated with a 3.6% decline in approval, compared to the pre-pandemic approval level. For example, a leader with a 50% approval rate before the start of the outbreak can expect a weekly decline in approval of 1.8 percentage points under these results. 

Figure 2 illustrates graphically the order of magnitude by splitting the sample into ‘high-case growth’ and ‘low-case growth’ countries. It plots the average approval of the two groups, including confidence intervals (shaded grey) from the week of the initial outbreak (100th confirmed case). The figure points to an initial increase in the average approval of both groups, consistent with a ‘rally-around-the-flag effect’ which boosts incumbents’ approval after major crises and disasters (Mueller 1970). 

For high-case growth countries, the initial increase in approval is smaller but still sizeable (roughly 10%). However, it quickly starts to decline again after the initial rally. After 13 weeks, high case growth countries are back at their pre-pandemic approval level. In contrast, governments in low case-growth countries do not see a fall in approval. After three months, their approval level is still 20% higher than their pre-pandemic level. This corresponds to a seven percentage point increase. The difference between the two groups is both quantitatively large and statistically significant. 

Figure 2 Government approval during the Covid-19 crisis, high- and low-case growth countries 

Note: This figure shows the percentage change in government approval on a weekly basis after the outbreak of the Covid-19 pandemic. The graph splits the sample into two groups: Countries with above median Covid19 case growth (“high case growth”) during the sample and countries with below median case growth (“low case growth”). The shaded grey areas show 90 percent confidence bands. The figure is based on an indexed sample, starting at the week of the 100th reported case in a given country. The data are smoothed using 3-week moving averages.

It should be noted that the correlation between case growth and approval is not independent of policymakers’ containment measures. In particular, the impact of case growth on approval is larger (becomes more negative) if governments did not implement tough policies to contain the infection. We measure the stringency of policies using the Oxford Covid-19 Government Response Tracker (Hale et al. 2020), which varies between zero (very loose) and 100 (very tough). Figure 3 depicts the effect of case growth on approval for different levels of stringency, including 90% confidence bands (dotted lines). The effect of case growth is significant only for values of policy stringency below 80 and increases as the policy stringency loosens. This lends support to the idea that the number of Covid-19 infections affect approval ratings primarily in countries which fail to implement tough policy measures to contain the spread of the virus.  

READ ALSO  Antitrust in the digital economy: IGM Forum survey

Figure 3 Interaction between case growth and stringency of government response  

Note: This figure shows the effect of Covid19 case growth on political approval for different levels of policy stringency. The dotted lines show 90 percent confidence bands.

The trade-off between health and the economy

In more recent months, some leaders were particularly keen on re-opening the economy at the potential cost of public health. To shed light on the ‘health versus the economy’ trade-off, we conduct a simple ‘horse race’ between health proxies (weekly Covid-19 case growth) and economic proxies (weekly changes in workplace visits and electricity usage). We further analyse these relationships over time, which allows us to compare periods of policy tightening (first weeks after the outbreak) with periods of policy loosing (second phase).

Figure 4 sheds light on the dynamics between approval and the main explanatory variables. The black line plots the level of correlation between Covid-19 case growth and changes in political approval over time. As can be seen, the correlation is positive early on and then becomes more negative over time. We compare this to the correlation coefficients relating to (1) changes in stringency (blue line) and (2) economic activity (workplace visits, red line). The correlation between approval and stringency is positive early on (as high as 0.35), but then falls to around zero after the 6th week. Changes in workplace visits are positively correlated with changes in approval, but the coefficients are small in magnitude. In comparison, the negative correlation between Covid-19 case growth and approval is more sizeable and has a stronger time trend. This suggests that the public becomes increasingly impatient with the increase in infections as time goes by. The economic performance, in contrast, is not a main predictor of approval, and this is confirmed in our regression analysis. 

Figure 4 Correlation of approval rates over time: health vs the economy

Note: This figure shows correlation coefficients (3-week moving averages) of changes in approval and the main explanatory variables, case growth, change in workplace visits (Google) and change in stringency. Correlation coefficients are computed based on the cross section of countries in a given period. The figure is based on an indexed sample, starting at the week of the 100th reported case in a given country. 

Why does government approval react  so strongly to changes in infections but not to changes in economic activity? We have no direct way to answer this question or to test why the public assigns so much weight on infection cases. One possible interpretation is preference-based, meaning that during a pandemic, the public cares most about health outcomes and less so about economic outcomes. This is also consistent with the finding that the public supports governments that take a tough policy stand. Another interpretation is that the public expects that the economy will not fare well until the pandemic is tamed. In this view, tough policies that bring down infections are a precondition for good economic outcomes in the medium and long run (in line with the evidence from the Spanish flu found in Correia et al. 2020). Indeed, there is growing evidence that individuals react to high infection numbers by restricting their movements, so looser policies do not necessarily imply more economic activity (Farboodi et al. 2020). A quick ‘reopening’ is far from guaranteed to result in a quick economic rebound.

READ ALSO  SE: AstraZeneca Vaccine Trial Likely Needs a Restart: Johns Hopkins

Concluding remarks

The political economy perspective of the pandemic remains poorly explored. Our analysis sheds light on the policy trade-offs that politicians face in a pandemic. In a nutshell, governments that placed more weight on health outcomes versus short-term economic outcomes gained political support. Moreover, this effect increases over time. At the initial stages of the pandemic, leaders benefit from a ‘rally around the flag’ effect and are granted the benefit of the doubt. But this ‘token of trust’ fades quickly. After about four weeks, growing case numbers increasingly hurt political approval, especially if no stringent policies were in place. We are only starting to understand the politics of pandemics, so it is possible that, looking ahead, the link between approval and health outcomes will change and/or that economic outcomes will matter more. 


Alvarez, F E, D Argente and F Lippi (2020), “A simple planning problem for COVID-19 lockdown”, Working Paper 26981, NBER.

Barro, R, J Ursua and J Wenig (2020), “Coronavirus meets the Great Influenza Pandemic”,, 20 March.  

Bol, D, M Giani, A Blais and P J Loewen (2020), “The effect of COVID-19 lockdowns on political support: Some good news for democracy?”, European Journal of Political Research, forthcoming.

Correia, S, S Luck and E Verner (2020), “Pandemics Depress the Economy, Public Health Interventions Do Not: Evidence from the 1918 Flu”, SSRN, 26 March. 

Eichenbaum, M, S Rebelo and M Trabant (2020), “The trade-off between economic and health outcomes of the COVID-19 epidemic”,, 20 July. 

Farboodi, M, G Jarosh and R Shimer (2020), “Internal and External Effects of Social Distancing in a Pandemic”, Covid Economics 9:22–58.

Giommoni, T and G Loumeau (2020), “Lockdown and voting behaviour: A natural experiment on postponed elections during the COVID-19 pandemic”, Covid Economics 41:69–99.

Hale, T, A Petherick, T Phillips and S Webster (2020), “Variation in government responses to COVID-19”, Oxford Blavatnik School of Government, Working Paper No. 31.

Herrera, H, M Konradt, G Ordoñez and C Trebesch (2020), “Corona Politics: The cost of mismanaging pandemics”, CEPR Covid Economics 50:3–32.

Herrera, H, G Ordoñez and C Trebesch (2020), “Political Booms, Financial Crises”, Journal of Political Economy 128 (2): 507–543.

Mueller, J E (1970), “Presidential popularity from Truman to Johnson”, The American Political Science Review 64 (1): 18–34.