Skip to main content

How partisanship and public health concerns affect individuals’ social mobility during COVID-19

Cite this dataset

Clinton, Joshua (2020). How partisanship and public health concerns affect individuals’ social mobility during COVID-19 [Dataset]. Dryad.


Rampant partisanship in the United States may be the largest obstacle to the reduced social mobility most experts see as critical to limiting the spread of the COVID-19 pandemic. Analyzing a total of just over 1.1 million responses collected daily between April 4th and September 10th reveals not only that partisanship is more important than public health concerns for explaining individuals’ willingness to stay-at-home and reduce social mobility, but also that the effect of partisanship has grown over time – especially among Republicans. All else equal, the relative importance of partisanship for the increasing (un)willingness of Republicans to stay-at-home highlights the challenge that politics poses for public health.


To obtain our data, we interview of 1,135,638 randomly selected respondents from the Survey Monkey platform between April 4, 2020 and September 29, 2020. These individuals are randomly sampled from the approximately 2 million individuals who take Survey Monkey surveys every day.

The data we analyze was collected by Survey Monkey using respondents who have consented to participate in the Survey Monkey audience panel and surveys. 

The anonymized data are weighted for age, race, sex, education, and geography using the Census Bureau’s American Community Survey to reflect the demographic composition of the United States. An additional smoothing parameter for political party identification based on aggregates of SurveyMonkey research surveys is included. Weights are generated each day using daily interviews Monday through Friday, and once over combined interviews on Saturday and Sunday. Party identification parameters are refreshed bi-weekly.

Data on COVID-19 prevalence comes from COVID-19 Data Repository by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University. These data report, on each day, the number of COVID-19 cases in each county in the United States.

As a first data processing step, we use county population estimates from the American Community Survey to transform county cases to be cases per 1000 residents.

We then calculate for each county, for each week, the change in the number of cases from the previous week in that county, using the median number of cases in each county for each week.

Consider, for example, interviewing an individual in zip-code 37206 (in Davidson county on Nashville’s East Side) on April 4th. On the week of April 4th in Davidson county (in which the zip-code is wholly contained) the median number of cases per 1000 was 1.001. In the week previous, the median number of cases per 1000 was .428. This individual, therefore, would be given the value .573 for change in county cases per 1000. This value is a fair representation of the fact that this individual, on this day, was witnessing a worsening public health situation in their immediate surrounding.

Survey Monkey respondents self-report their zip-code. To assign respondents to a county we determine the county in which the majority of the residents of that zip-code live in. For 86% of respondents, their zip-code is entirely contained within one county.

Finally, we merge the data on how the number of COVID-19 cases is changing in each county on each day to the Survey Monkey data.

Usage notes

Data was analyzed in R.  All code needed to replicate tables and figures are included.