How you vote affects your health

The Civics & Covid-19 project team maps State colors based on voter choices in the 2008, 2012, 2016, and 2020 presidential elections. In November 2020, over 81 million USA citizens voted for President Biden. On January 20, 2021 when President Biden took office, State colors were muddy because more and more cases of Covid-19 were reported every day. He promised to stop the spread of Covid-19 by giving 100 million vaccinations in the first 100 days of his presidency. Before the April 30 deadline, he reported that 200 million COVID-19 vaccinations had been given, double his initial goal.[1] Now, our colors are brighter and our health is improving.

During Biden’s first 100 days, we observed some interesting reporting trends. Every day we download cumulative raw data from Johns Hopkins flat files.[2] We compute daily statistics and store them in a MySQL database that powers our online reports. Our maps and reports show Covid-19 data categorized by USA voter choice. For example, Californians voted for democrats so the State of California is blue and California appears at the top of the blue table below. Texans voted for republicans so the State of Texas is red and Texas appears at the top of the red table below.

Covid-19 from April 12, 2020 through April 30, 2021

Has anyone really recovered from Covid-19?

Look at the “Recovered” columns in all of the tables. Notice that all the numbers in the Recovered columns are zero. This is new. In 2020, States reported how many people recovered. Red States reported more people recovering than blue or purple States (56% red and 71% light red, as opposed to 14% blue, 48% light blue, and 34% purple).[2] But Covid-19 recoveries are difficult to monitor because Covid-19 is a new virus. Scientists don’t know what long-term harm it may cause. According to the Mayo Clinic, more research is needed to assess the long-term effects of Covid-19, such as lasting heart damage, scarring of the lungs, brain damage that can cause strokes and seizures, and an increased likelihood of blood clots.[3] Reporting zero recoveries better reflects this uncertainty.

Do Covid-19 statistics tell the truth?

Using raw data, we compute a simple linear regression for each State, then rank the States accordingly. The color of each State is muddy or bright based on how it ranks among all the other States. When all the States report data consistently then data scientists can create good models and make predictions about how Covid-19 will affect us. If our elected leaders do not prioritize consistent reporting then the data may not be consistent enough to develop an accurate model. Thus, our simple ranking of raw data reveals a candid observation of how some of the people we elected are prioritizing our health.

Take a look at Confirmed Cases of Covid-19 in Michigan. The Michigan data tells a story about what happened with Covid-19 reporting over the last four months. In the beginning, Michigan was reporting about 2,000 cases daily. Like most stories there was a problem in the middle but it gets resolved by the end. The problem occurred because more and more cases were reported until they reached a peak of more than 10,000 cases daily. In the end, the number of cases started going down giving Michigan a story with a happy ending.

Now, take a look at the charts during the same time period for Iowa, Missouri, and New Jersey. The curves are almost too small to see and there are spikes above the line and below the line. We noticed these charts because they caused all the States on our map to turn unexpected colors. One day, all the States were black except for Missouri. A few days later, all the States were brightly colored and Missouri was black. That prompted us to investigate.

The story told by this Missouri chart is that data was not reported consistently. Around day 70, Missouri reported 69,778 cases on a single day! Then a few days later subtracted 8,492 cases. The Iowa and New Jersey charts tell the same story. Around day 50 Iowa reported 27,302 cases then a few days later they subtracted 27,741. Similarly, New Jersey reported 61,850 cases then subtracted 9,005.

Unfortunately, this type of reporting makes it difficult to see trends in the data, not only, for these States but for the entire map. Honest mistakes can and do occur butdata experts know how to smooth the data to show what would have happened if the mistake had not occurred. Knowing the importance of reporting data consistently, what do these spikes tell us about the priorities of the elected leaders in these States? Did they properly fund a Covid-19 reporting system along with data experts to make sure their data was reported consistently?

Other data reporting problems have been investigated by experts such as those working at the Institute for Health Metrics and Evaluation (IHME). They recently published results of a study explaining that variations in Covid-19 reporting and other factors resulted in deaths due to Covid-19 being under-reported.[4] They compared the number of expected number of deaths in 2020 to the actual number of deaths reported. They took into account six factors contributing to deaths in 2020. Some Covid-19 deaths were not properly classified because Covid-19 testing was in short supply. There was a shortage of healthcare workers and some people died because they could not see a doctor. Anxiety and depression increased during the pandemic along with the suicide rate. Because people were staying home during the pandemic, there were fewer deaths from accidents and some infectious diseases, along with a decline in deaths from chronic conditions such as chronic respiratory disease. The results of this study indicate that there were more than 900,000 deaths in the USA.[5]

Looking forward to 2022

When data are reported consistently, then data experts can show us trends and predicted outcomes. Elected leaders can rely on data to make decisions that will help improve our health. A good example of consistent reporting during President Biden’s first 100 days was that all States reported zero recoveries. This reflects the fact that scientists do not know the long term affects of Covid-19 on our health. As we saw in the blue, purple, and red States of New Jersey, Iowa, and Missouri, when elected leaders do not place a high priority on consistent and accurate reporting then we lose confidence that their data shows us what really happened. Inconsistent data reporting led to under-reporting deaths according to the Institute of Health Metrics and Evaluation.

The Civics & Covid-19 team will keep track and report observations in January 2022.

Click here to read what happened in 2020:

https://cy-nelson.medium.com/how-you-vote-may-affect-your-health-897ffcc96095

References

[1] https://www.npr.org/2021/04/21/989487650/biden-says-goal-of-200-million-covid-19-vaccinations-in-100-days-has-been-met

[2] COVID-19 Data Repository by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University. 2021. https://github.com/CSSEGISandData/COVID-19

[3] “Why aren’t coronavirus recoveries always reported?” By Catherine Marfin, Dallas Morning News, October 2020. https://www.dallasnews.com/news/2020/10/08/how-does-anyone-really-know-when-someone-has-recovered-from-covid-19/

[4] https://www.npr.org/sections/coronavirus-live-updates/2021/05/06/994287048/new-study-estimates-more-than-900-000-people-have-died-of-covid-19-in-u-s

[5] http://www.healthdata.org/special-analysis/estimation-excess-mortality-due-covid-19-and-scalars-reported-covid-19-deaths

Researcher, technologist, and paralegal student