The Government and COVID-19: Anger, Policies and Partisanship

Laura Biester; The Pandemic Project, University of Michigan

The voices of government officials have been a constant presence during the COVID-19 pandemic, and unsurprisingly, most people have something to say about what their elected officials are doing. A sample anonymous responses that mention the government are listed below:

I feel that my government has not honestly effectively dealt with what they knew was coming and we are totally unprepared.

I wasn’t anxious I was confident in the plans that were being rolled out by my local government and my work.

I am frustrated that our government did not act quickly enough at the beginning of the outbreak to prevent or lessen the impact of the pandemic in our country.

In this blog post, we seek to gain a greater understanding of people’s feelings about the government response to the pandemic, and factors that may influence those feelings, including partisanship and lockdowns. We use responses to the COVID-19 survey to measure how people feel about the government at different levels (e.g. federal, state), and the ways in which those around them are responding to the situation that has upended our lives. In this analysis, we will focus on respondents from the United States, who make up more than 75% of overall respondents. We find that:

  • There is a strong tendency among our respondents to be displeased with the government; satisfaction is higher among conservatives than liberals.
  • Liberals express positive emotions towards their state governments, while conservatives express positive emotions towards President Trump.
  • Government lockdown orders have a clear effect on how people feel about the reaction to the crisis.

Who is pleased with the government: the partisan split

In one version of our survey, we asked respondents to state, on a scale from “not at all” to “a great deal,” how satisfied they were with how their government has handled the crisis. More than 50% of participants reported that they were extremely dissatisfied with how the government had handled the crisis. Among all of the questions we asked, the most strongly correlated with government response was one’s political orientation (Spearman’s r = 0.40), meaning that those who reported being more conservative were more satisfied with the government’s response. The graph below shows the percentage of respondents who report high and low satisfaction with the government response; what we find is that while the entire population seems somewhat dissatisfied (fewer than 10% in each group reported the highest satisfaction level), the dissatisfaction levels drop as respondents become more conservative.

Percent of respondents reporting satisfaction with government (by political orientation)

The partisan nature of this response is likely a reaction to how the White House has handled the pandemic. However, this does not capture the full picture; we note that there are multiple levels of government that people may form opinions on. Somewhat shockingly, we see that nobody who claims to be “extremely liberal” indicates any level of satisfaction with the government, despite the fact that many respondents are from states with liberal governors (e.g., California, ~10% of respondents). A possible interpretation is that when Americans are asked about the government response, they will only report satisfaction if they are satisfied with all levels of government; even if they are liberal and living in a state with relatively strict restrictions, such as California, they will still think of their discontent with the White House when responding. While the liberal response appears to be more focused on the federal government, the conservative response is mixed; although there may be some unease among conservatives about the federal government’s pandemic response, they may also be considering their displeasure with state officials when answering this question. In order to tease out the exact cause of these responses, we suggest that future work asks about government orders on a scale from too strict to too lenient for multiple levels of government: local, state, and federal.

Given our question about the government, we still need to confirm whether people are dissatisfied because they think that too much is being done, or because too little is being done. We therefore define the metric perception of COVID response based on survey questions that assess how people feel about the response of others in relation to COVID-19; on one side of the spectrum, we have people who believe that others are making too big a deal about COVID-19 (too strict), and on the other side, we have people who believe that those around them are not taking the situation seriously enough (too lenient). We find among those who are dissatisfied with the government response, liberals consistently believe that the response is too lenient; however, among conservatives, the responses are mixed (just like government satisfaction). This supports the idea that some conservatives who are upset may be upset with harsh restrictions, while some believe that others are being careless about the pandemic and would prefer harsher restrictions.

Perception of COVID response among respondents who indicated that they were dissatisfied with the government reaction. We codified “extremely conservative” and “conservative” as conservative and “liberal” and “very liberal” as liberal. Those who identify in the middle are not included in this analysis.

We find other ways in which this metric is related to the government’s response to COVID-19: people felt that the response was stricter in mid-April, when many states had state-wide lockdowns. However, we saw people again begin to feel that the response was too lenient when states began to lift lockdowns. This suggests that what the government does has a serious effect on people and how secure they feel in their communities.

The effect of lockdowns on perception of COVID response. We find that the number of states that have lockdowns is correlated with people’s perception of the response (Spearman’s r = .74). Values for perception of COVID response are seven-day rolling means with missing values linearly interpolated. Both series are mean normalized to appear on the same axis.

We believe that perception of COVID response is intrinsically linked to the values of individualism and collectivism; those who perceive the response as too lenient do so under the collectivist belief that others should be taking actions (staying at home, wearing masks) that negatively impact their lives for the good of society as a whole. Those who lack this perception have indicated that others are making too big a deal of COVID-19, signaling their more individualist desire to return to their previous lifestyle.

The fact that the most severe pandemic in one hundred years is coinciding with a presidential election has led to the politicization of the virus largely along these two ideals, with Republicans pushing for individual liberties while Democrats push for policies that they believe will save more lives in the long run, with an immediate cost. People tend to be unhappy with the government response to the pandemic, and its status as a political talking point cannot be ignored when considering that dissatisfaction; we now turn to looking at the specific emotions that people portray when discussing the government to examine how the government’s stance on the pandemic affects people.

Which emotions are associated with the government during COVID-19?

We examine the free-text responses in order to understand the emotions that people portray when they discuss the government. We created lists of words related to government1 in multiple ways:

CategoryExample Words
General Governmentgovernment, govt
President TrumpTrump, Donald, President
State Governors and LegislatureGovernor, Cuomo, Abbott
Partisan Politicsconservative, democrat, wing
Miscellaneousstate, CCP, propaganda
No Government MentionDoes not mention words in any government category

We identify sentences that mention these government categories and use the LIWC categories for positive emotion and three negative emotions (anger, anxiety, and sadness) to understand the emotions of our respondents towards government. Results vary drastically when considering those with different political orientations, so we carry out the analysis for both liberals and conservatives.

Average percentage of emotion words in each government category among those who identify as conservative or extremely conservative.
Average percentage of emotion words in each government category among those who identify as liberal or extremely liberal.
  • This analysis supports our hypothesis that liberals are generally angry at the President, and comparatively pleased with their state legislature. Another finding that is consistent with our earlier analysis is that liberals tend to express far more anger when speaking about the government than conservatives.
  • Conservatives tend to express positive emotions when writing about the President, and are comparatively less positive when speaking about their state government. When considering negative emotions, it is interesting that the only category in which the emotions about the state government exceed those about Trump is anxiety.
  • We tend to see that negative emotions among liberals have a clear focal point: the President, and to a lesser extent, “the government” more generally. Meanwhile, conservatives express more anger towards the government generally and partisan politics; the conservative responses falling under partisan politics tend to be directed towards the other side (66% of responses from conservatives mentioning one party mention Democrats).
  • More generally, responses that do not focus on government tend to express more sadness and anxiety, while responses that focus on government express more anger.

Looking Ahead

It has now been four months since the WHO declared COVID-19 to be a pandemic, and we will continue to face debates about the competing values of individualism and collectivism. It is very likely that these debates and the policies that trigger them will continue to cause anger. While partisan politics and the upcoming election threaten to further divide the country, spurring on COVID-19 related debates on public health initiatives like widespread mask usage, we hope that as a country, we can settle into a new, temporary status-quo that will minimize deaths and be accepting of our humanity.


  1. The complete lists of words used to identify different levels/aspects of government:
    General Government: government, govt
    Partisan Politics: conservative, democrat, democratic, liberal, republican, wing
    President Trump: Donald, President, Trump
    State Governors and Legislature: Governor, legislature, all governor’s names excluding Baker, Brown, Deal, Justice, Little, and Wolf (more likely to be associated with other meanings)
    Additional words: protest, propaganda, congress, senate, China, Chinese, CCP, state, country, federal, party