In today’s post, we are going to take a look at various forecast models and possible electoral map outcomes and discuss what each could mean for the 2020 election. Many swing states have been rated differently by each forecaster, so it will be interesting to see how various states are rated. As many forecasts are updated numerous times a day, all characterizations are correct as of writing.
FiveThirtyEight Projection
The FiveThirtyEight Projection is one of the most reliable and trustworthy projections. They account for bias in pollsters and also make use of demographic trends to predict each state. Margin colors used in the map below are based on how likely each party will carry each state: >90% for safe, >80% for likely, >60% for lean, and >50% for tilt.
I agree with all of these characterizations except for Florida and Nebraska’s Second District. These should definitely be quite likely to go to Joe Biden, based on polling data and history. Florida, especially, has been showing consistent leads for Joe Biden in reliable polls since April.
Notably, FiveThirtyEight is often very conservative (as in cautionary) with their estimates, and they were also one of the few models on election day in 2016 that still gave Donald Trump a 30% chance of victory.
JHK Forecast
The JHK Forecast is another commonly-cited forecast. It has been quite reliable in the past, and they also have projections for the House and Senate elections, too. Margin colors used are the same as those used for the FiveThirtyEight Projection above.
It is interesting to see how much a map can differ between forecasters. Because JHK is not as cautionary as FiveThirtyEight, it has moved Colorado and Maine At-large from likely to safe Democrat; Pennsylvania, Wisconsin, and Minnesota from lean to likely Democrat; Florida from tilt to lean Democrat; Maine’s Second District, Iowa, and Ohio from lean to tilt Republican; but it has also moved South Carolina from likely to safe Republican.
Clearly, JHK’s model is more favorable to the Democrats than FiveThirtyEight’s.
The Economist Forecast
The Economist also has a prediction model which is very highly regarded too. Notably, they only give statewide ratings and don’t provide separate data for Nebraska and Maine’s districts, so they are both considered safe here. Margin colors used are the same as above.
This map is pretty similar to JHK’s map. The only differences are that Nevada and Pennsylvania are lean rather than likely Democratic and Iowa and Ohio are lean rather than tilt Republican. In general, it doesn’t differ much from the other maps.
RealClear Politics No Tossups Map
The RealClear Politics (RCP) normal election map is notorious for classifying most competitive races as tossups (even races like Missouri, New Hampshire, and Nevada), so whenever RCP’s data is used, it’s always the no tossups version. However, the no tossups map does not include margins or probability, so states are just characterized based on party. RCP is well known for having averages for every state, but its weighting is worse than FiveThirtyEight because it considers a poll from a bad pollster equal as one from a good pollster.
This one is interesting because Ohio flips blue, whereas none of the other maps characterize it as such. The other states are all characterized similarly to the other projections above.
Polling Average Map
This map is based on polling averages from FiveThirtyEight. Where there are fewer than two polls available in any given state, 2016 results are applied. I will be making a more detailed post on this in the future. Margin colors are winning margins, as above.
This map seems pretty accurate compared to most predictions. It may have overestimated Democrats in Maine’s Second District slightly, but for the most part, it seems like just about any other projection.
Best-Case Scenarios
Here are the maximum possible scenarios for Joe Biden and Donald Trump, assuming each candidate won a landslide election. These are “realistic,” meaning that it would be theoretically possible based on polling data. None of these are likely to happen, but it is fun to consider. Margin colors are based on winning margins (>15% safe, >5% likely, >1% lean, >0-1% tilt).
Joe Biden landslide:
This is the scenario in which Joe Biden wins all the swing states, and other traditionally safe Republican races narrow up. He gets 413 votes, the theoretical cap for Joe Biden in the Electoral College.
Donald Trump landslide:
You can really see how dire things are for Trump just by looking at this map. He just about narrowly manages to keep Pennsylvania and Arizona, the only two swing states that could realistically go to him. Michigan, Wisconsin, and Minnesota are basically gone for him at this point, as the polling average have Biden up six to seven points (in 2016, the margins were way narrower in these states).
Newshacker Projection
The following map is my projection, aggregated from all sources above, and adjusted slightly by me. Margin colors used are based on how likely each candidate will win each state, as per the ones used for the landslide scenarios (>90% for safe, >80% for likely, >60% for lean, and >50% for tilt).
And here is the map based on the expected winning margins (>15% safe, >5% likely, >1% lean, >0-1% tilt):
For a detailed explanation of the Newshacker Projection, see the post here.
I hope that these maps gave you a better insight into the 2020 election as to how things could go, and how pollsters are rating races. Stay tuned as my analysis on the House elections are coming soon. Make sure to subscribe so you don’t miss it.
Polling data is correct as of writing. Polling data may have changed slightly since writing. Polling averages are based on the FiveThirtyEight average, and their website provides the most up-to-date information.