*Disclaimer: This article was written before the President tested positive for COVID-19.
In 2016, state polling in Wisconsin resulted in one of the biggest misses leading into what many considered an Election Day surprise when Donald Trump defeated Hillary Clinton both in Wisconsin and nationally to win the election. Not a single poll tracked on Real Clear Politics (RCP), a site that provides polling aggregate information, showed Trump winning Wisconsin. In fact, every poll from June through November showed Hillary with at minimum a 3-point lead and the average going into Election Day showed her with a 6.5 point lead. In 2016 FiveThirtyEight gave Hillary an 83.5 percent chance of winning Wisconsin going into Election Day, and now in 2020 it has Biden with an almost identical 83 percent chance of winning. FiveThirtyEight would note that this means Trump has nearly a 1-5 chance of winning, which is about the same percentage chance as a Hamilton lobbyist making a 3-point shot. Not likely, but certainly not inconceivable.
What should be made of the similarities of 2016/2020 polling that shows a parallel pattern for Joe Biden versus President Trump in Wisconsin? Does it foretell the same result happening or is it different this time around?
No one knows for sure, but there are some differences. The first difference between 2016 polling and 2020 polling is the sheer volume of state polling being done, both in Wisconsin and across the country. In 2016 there are only two polls on RCP for the entire month of September. In 2020 there are 13 Wisconsin polls. Given that these are statistical models, it does provide a little more of a foundation in what the current polling average is (Biden +5.5).
A second difference comes in the number of undecideds. In 2016 there appeared to be nearly double the amount of undecideds going into Election Day. With Trump winning those late breaking undecideds by a sizeable margin, the lead Hillary had was surmountable. Especially when Democratic turnout lagged in Wisconsin overall. With what appears to be less undecideds available going into Election Day, and the unlikelihood that these undecideds would break in the same fashion as they did in 2016, Trump seems to have a narrower path to winning Wisconsin in 2020 versus in 2016.
2020 brings several elements into play that could have impacts on the results, results we may not fully know until days after election day. There will be far more mail-in ballots than ever before. Will all the mail-in ballots be counted or will legal challenges to their validity be successful? Will there be other problems with counting millions of mail-in ballots? Also, Election Day voting could see an impact from COVID-19. If certain areas are seeing spikes in the virus could that lead to lower voter turnout in those areas? Will there be an October surprise? What if that surprise is a successful vaccine? These are all questions that could possibly upend what statistical models see as a probable Biden victory in Wisconsin.
Given all of that, Biden continue to maintain a 5ish point lead in Wisconsin with less than 5-weeks to go and if given the choice, President Trump would trade positions with him in a heartbeat.