The Risk To Not Delaying Primary Elections

14 states across the country have delayed primary elections into the summer. But many have not. What, if any, is the cost of not delaying these elections during the COVID-19 World Pandemic? We review what the numbers say.


Unless you’ve been living under a rock you’ve taken note of the WHO declaration of COVID-19 as a world pandemic on March 11th. On March 13th President Trump declared the COVID-19 a national emergency. Before these announcements, the U.S. was well on the path to determining the next Democratic candidate for president by way of electoral primaries. As of March 17th, 28 states and territories have held primaries. As of March 28th, 14 states have delayed their primaries. These states include Alaska, Connecticut, Delaware, Georgia, Hawaii, Indiana, Kentucky, Louisiana, Maryland, New York, Ohio, Pennsylvania, Rhode Island, Wyoming, and Puerto Rico have rescheduled their contests. But should these states have delayed their primaries? Should remaining states that have not delayed primaries move forward? What is the cost of additional illnesses, if any, that going through with this action may lead to? We crunch the numbers below, but please take note, the information is quickly changing, and these numbers are as of March 28th, 2020.

Virus Numbers

How contagious is the virus? Utilizing social distancing most models indicate that one person, under strict social distancing (shelter in place or similar tactics) spreads the disease from one person to two and a half people within 30 days. Many states have employed this tactic and our modeling will account for this. Without social distancing, the number vastly increases to 400, or 160 times more people in 30 days. Individuals that contract the disease are expected to need hospital care 15% of the time, ICU care (utilizing ventilators) 5% of the time and can be expected to pass away 1% (based on Dr. Fauci’s estimate) of the time.

Election Numbers

Election numbers vary wildly from state to state, especially when dealing with primary elections or elections that are not presidential. This requires us to look at a specific example. In our example, we will evaluate one of the states that have not delayed their primary elections and are expected to heavily influence the presidential election, Wisconsin. 

Wisconsin’s primary voter turnout is 14.5% of a voter eligible population of 4.4 million. This estimates that roughly 650,000 people will vote in the upcoming primary election. Of these voters, 25% will vote absentee. This leaves 486,000 voters turning up to the polls.

COVID-19 Positive Voters

Wisconsin has seen 1,112 cases to date. Models estimate that the U.S. has captured 12% of the COVID-19 cases, leaving 88% of cases to go undetected. Using these metrics Wisconsin is likely to have 8,155 unknown or asymptomatic cases of COVID-19, some of which will turn up at the polls. With 11% of Wisconsin eligible voters going to the polls, this estimates that roughly 887 in-person voters will be COVID-19 positive. In 60 days, this would net 2,217 new cases of COVID-19 under strict social distancing, which Wisconsin Gov. Evers has implemented through April 24th.

The COVID-19 is expected to peak in Wisconsin around May 22nd or roughly 60 days from now. Under strict social distancing, we can expect every single case to result in 6.25 cases in 60 days.

The Cost of Continuing with Elections

Simply becoming ill however is not the cost of holding elections. Individuals become ill frequently and we don’t delay elections for this. The fear of the COVID-19 pandemic isn’t necessarily becoming ill but rather needing hospitalization or even death.

The virus is expected to require hospitalization for 15% of those that are positive for the virus. In Wisconsin, this would require 831 new hospitalizations. At any given time, Wisconsin has approximately 7,000 open beds and the addition of these 831 new hospitalizations increases usage by approximately 12%.

Also, the virus is expected to require 5% of cases to require the usage of ventilators. This would translate to an additional 277 ventilators required. No concrete numbers have been published by Wisconsin as to how many ventilators are available, but the national average is 19.7 per 100,000 residents. For Wisconsin, this means roughly 4,500 ventilators or 2,600 available when you apply the open bed factor.

Finally, the worst-case outcome for acquiring COVID-19 is death. Using a 1% death rate this estimates that 55 Wisconsin residents will die as a consequence of coming into contact with another COVID-19 person due to in-person voting. 

The Effect of a 30-Day Delay

If Wisconsin delays the election by 30 days, we might see a dramatic decrease in most of the numbers above. This is largely due to the reduction in transmission. The virus spreads to two and a half people every 30 days. Because of the exponential effect, this means that six and a quarter of people will contract the virus in 60 days. All else being equal it would be a moot point to delay the election, you’d simply be delaying the illness. In this case, all else is not equal. The virus is expected to peak in 60 days meaning that after those 60 days the rate of infection will decrease rather than increase. Any delay that pushed the contraction of the virus closer to that peak from now (pre-peak) is beneficial.