Tag: 2020 Presidential Election

Presidential Race Ratings Update: Michigan, Nebraska’s Second Congressional District, and Virginia

Another week, another series of ratings changes. We’re three weeks away from Election Day and looking every day at the wave of new polls and information coming in. After playing it safe for the last few weeks, we’ve come to realize that there are a couple states and one congressional district that are clearly favoring one candidate over another as the race continues to favor Biden, so we’re adjusting our ratings for them accordingly. 

Michigan

Over the course of the last month, Biden has routinely received high-single digit or double digit polling leads in the Great Lakes State. Of the “Blue Wall” states that flipped to Trump, Michigan was the narrowest and most surprising, going for him by just under 11,000 votes. Last cycle, in the home stretch towards Election Day, Clinton led in the RealClearPolitics voting average of the state by 3.6%, and Michigan was largely considered a “Lean Clinton” race. This year, Biden leads by 7%, is in much better shape with white voters than Clinton was, and Democrats are looking to strengthen their position in a state where the “blue wave” was felt strongly in the 2018 midterms. Biden doesn’t just have an advantage in this race, he simply looks likely to outright win Michigan.

Nebraska’s Second Congressional District

I’m so glad we get to do a ratings change here because it lets us explain a quirk of the Electoral College. Maine and Nebraska both allocate electoral votes based on congressional district (so, whichever candidate wins the vote in that district, wins one elector) and then its two at-large electoral votes to the winner of the state overall. This is generally considered a bad system, perhaps even worse than the winner-take-all system most states use, as it would make gerrymandering far more dangerous and contentious than it already is, and would have delivered an electoral inversion (wherein the popular vote winner loses the presidency) not once in 2000, not twice in 2016, but three times this century, giving Mitt Romney the presidency in 2012. 

Getting back to the point, we hedged on whether Nebraska’s Second Congressional District was less competitive than a pure toss-up. Obama won the district in 2008, and then he and Clinton both lost it in the succeeding elections, but Biden has led in every poll of the district so far. The district is characteristic of the shift that has happened over the last few years under the Trump presidency. It is a mix of urban and suburban, covering the city of Omaha, and with a competitive congressional election there to boot, Biden is at least the favorite in the “Big O.”

Virginia

We talked about how Virginia is not as exciting as it used to be in our Senate preview for the state, and it’s true. Though it used to be a Republican stronghold, over the past twenty years it went from being a toss-up, to a lean Democratic state, to a likely Democratic state, and now has seemed to have completed the transformation into a safe Democratic state. Biden has never been behind Trump in a poll of the state, and the campaigns and state itself have acknowledged its time to shine as a battleground has passed. It’s a state whose center of economic and political power increasingly revolves around the very-liberal DC beltway and we believe 2020 is the first year that we can truly say Virginia is safe for the Democratic candidate. When you look back on political history in this century, it’s fair to say that no state has moved so dramatically from solidly Republican to solidly Democratic than Virginia, and that is in large part the fault of Donald Trump.

Presidential Race Ratings Update: Georgia, Ohio, and Utah

In the month since we released our 2020 presidential election ratings, we’ve been forced to reckon with two major news events for which the full impacts on the presidential election are unclear, but certainly relevant: the death of Justice Ruth Bader Ginsburg, and President Trump’s contraction of COVID-19. While we at The Postrider have generally held to the philosophy that this race was largely decided months — if not years — ago, considering how few undecided voters there are and the intensifying levels of partisanship. We also expected (which, in hindsight, feels stupid) President Trump to behave more like your traditional incumbent and to try and consolidate support across the spectrum rather than rely on his enthusiastic but minoritarian base.

When we released our presidential ratings, we noted that we would offer a ratings-focused take on the presidential race at large, and we’ve been keeping a watchful eye on the situation in a much larger group of states than likely anyone thought they’d be looking at four years ago. And, in a situation where we decide a state’s rating truly merits adjustment, not just in the short term, but for Election Day itself, we would provide a justification and background as to why and update our presidential map to reflect that. This week, we will be making our first three ratings adjustments by moving both Georgia and Ohio from “Lean Trump” to “Toss-Up” and moving Utah from “Likely Trump” to “Safe Trump.”

Georgia

Unlike other “Lean Trump” states, which include Texas and Iowa, Trump’s lead has actually decreased in the last month in Georgia, and Biden is now ahead or even in many recent polls. Considering the polling looks similar to what we’re seeing for Biden in North Carolina now, those states’ similar demographics (though Georgia has a much larger share of African-Americans, which should help Biden further), and the narrowing of Georgia’s concurrent Senate races, we see no reason to pretend Georgia is less competitive than it ought to be, and that means Biden has about the same odds to win it as Trump does.

Ohio

After several high quality polls have put Biden even or ahead in the Buckeye State, it’s getting harder to ignore the signs that Ohio may be reverting to the mean after a big swing towards Trump in 2016. Democrats and the Biden campaign have also stepped up their efforts there, realizing that, if Trump cannot carry Ohio (or Florida) the election is likely sealed. Trump is behind with educated white voters and in the suburbs, and his coalition looks weaker this time than last, especially as neighboring states with similar demographics and economies, like Pennsylvania and Michigan, pull away from his reach. With that in mind, we are adjusting our rating accordingly. Ohio is now even odds for Trump and Biden.

Utah

Whatever we were thinking might happen in Utah — a Mitt Romney endorsement of Biden, a further degradation of Trump’s standing with Latter-day Saints, an expansion of Biden’s lead in other states with large Latter-day Saints populations like Arizona and Nevada — hasn’t panned out. Biden has not polled higher than 44% in any credible poll of the state, and with no third party “McMuffin” spoiler this year, Trump looks soft, but for this year, safe, in Utah. 

The Postrider’s 2020 Presidential Ratings

Can you believe general election season is just getting started? Yes, with the conclusion of the Democratic and Republican national conventions in August, we’re officially off to the races! Call this election whatever you want: the “most important election in the history of our country”? Why not. A “battle for the soul of this nation”? It sure seems that way! The coronavirus election, a referendum on the president, the “George Floyd election”, or a skirmish in the incessant culture war; let’s just go with all of the above. No matter how you see it or how you come down, we hope you’re as excited as we are, because 2020 is the first year The Postrider will have the privilege to unveil our own ratings for a presidential election, providing background as we keep an unwavering eye on an election that may go long past Election Day.

You may recall our 2018 Senate race ratings and analysis, which served as a predecessor and dry run for the big showdown this year. We are planning to unveil our 2020 Senate ratings in the coming weeks as well, which will come later because some states hold late Senate primaries (Delaware’s primary isn’t until September 15, less than 50 days from Election Day!), but our presidency ratings will be different from those as well: broader, bigger, and less personal. So, here are some key things to keep in mind about our process this cycle:

Ratings-Focused

In our 2018 Senate ratings, we focused on telling the individual stories behind the candidates — because candidates matter more in localized races, and less so on the presidential level — and every state’s Senate race is different, with different personalities and dynamics. The presidential election is, by definition, nationalized and consistent. It’s the same two candidates across every state, so that kind of candidate-focused narrative is less useful. That’s why we’re choosing to focus on ratings instead of that individual analysis. 

Each state will be rated on a scale as follows:

“Safe”
The candidate has over a 99% likelihood of winning the state. This is an all but assured state for that candidate, but be mindful that if you were to run the election hundreds of times, thus with hundreds of “Safe” states, you’d expect less than one out of 100 to flip; that’s very low and very rare, but it does happen. Nonetheless, “Safe” means we are confident in a candidate’s odds in a given safe state.
“Likely”
The candidate has at least an 85% certainty of winning the state, so the other candidate has at least an outside chance of winning the state in 1% to 15% of cases. In our map, we have around ten “Likely” states, so if one of them flips; that would not be surprising. It would also not be totally crazy if none of them flip, though the more there are, the more you should expect one to flip.
“Lean”
The candidate is slightly favored but is by no means safe; the other candidate has somewhere between a 15% and 40% chance of winning the state too. In our map, we have just under ten of these states, so it should actually be surprising if at least one of these states does not flip.
“Toss-Up”
The race in this state does not clearly favor either candidate over the other. Each candidate has between a 40% and 60% chance of winning the state.

We determine ratings based on polling data, demographic data, news events, historical trends, the electoral environment in similar states,1Some states are highly correlated, like the midwestern states of Michigan, Pennsylvania, Wisconsin, Ohio, and Minnesota. If things take a dramatic turn in Wisconsin, you’d expect similar movements across these states based on their similar demographics, political histories, urbanization, population size, economic similarities, etc. Note that geography is not always the best determinator for these variables; for instance, Washington state and Colorado are also similar; as are Maine and Montana. how states are responding to COVID-19 and planning their election procedures, and all sorts of other smaller variables. It’s slightly more scientific than a “gut instinct” but less so than an explicit statistical model. We’re aiming for something closer to the Cook Political Report, Sabato’s Crystal Ball, or Inside Elections,2Inside Elections’ publisher is also employed by this author’s employer and The Postrider is completely independent. I mention this for full disclosure, Washington is a small town. with the intent to handicap the race, more so than our idols at FiveThirtyEight, though our various ratings are generally similar other than at the margins. All of our ratings are approved and determined by both the Editor-in-Chief, Michael Lovito, and yours truly, the State & Science Editor.

Update-Oriented

If we decide a state rating merits adjustment, we will publish that under the hover-over for a given state, and write a larger article if necessary. For example, at the time of publication, we have Colorado rated as “Likely Biden”, but if a few weeks from now Trump has improved his numbers dramatically amongst white college-educated voters and Latinos, we’d be inclined to revise our rating in Colorado (which has large shares of both kinds voters) down to “Lean Biden”. We will provide a ratings’ history when you hover over any given state which will include the date of the ratings revision(s) and an explanatory note or link to a larger piece justifying the change. This will make interacting with our ratings map a little more interesting, involved, and useful to watch as the race goes on and provide more context where necessary.3And give us a better starting point when we work on our 2024 ratings for the next cycle! 

All ratings revisions will be approved by both of the editors on this project, and based on the same criterion noted above in terms of changes in polling, how the candidates are doing amongst certain demographics, state-by-state correlations, etc. We won’t change ratings if we believe they are likely to change by Election Day, since that would defeat the purpose of forward-rating these races, but we will change as often as we need to if compelling evidence points to better or worse prospects for the candidates in a given state.

The Overall Projection

Our projection of the total electoral votes per candidate is not a prediction, it is a mathematical allocation based on the candidate’s odds for given electoral votes. For example, we have Tennessee as “Safe Trump” , and with 11 electoral votes, we multiply 11 by 0.99 (as “Safe” we’ve determined means it’s at least better than 99/100 odds that Trump wins that state), or 10.89 votes and 0.11 for Biden. Meanwhile we’ve rated Virginia as “Likely Biden”, so we multiply its 13 electoral votes by 0.85 giving us 11.05 votes for Biden and 1.95 for Trump. “Lean” is multiplied by 0.6 and “Toss-Up” by 0.5. Once all states’ electoral votes have been split in this way, we round down to the nearest electoral vote. 

Maine and Nebraska split their electoral votes by congressional district, so we’ve accordingly rated the state at large (two electoral votes) and each of their congressional districts and applied those ratings to their individual and statewide votes. These states are shaded to reflect the diaspora of ratings at large and for the individual districts.

At the time of publication, our projection has Biden at 315 electoral votes, 45 more than the 270th vote he needs to win, and a slightly more conservative estimate than FiveThirtyEight’s or The Economists forecast models.

Remember that we’re not making predictions, we’re just providing an educated assessment on the state of the race in every state.4And the District of Columbia! So don’t write us a mean note after Election Day saying “you had Wisconsin as ‘Lean Biden’ but then Trump won it by one percent, idiot!” No, we didn’t predict that Wisconsin will “only” go for Biden by a little, we’re merely stating it’s more likely than not to go for Biden, but Trump still absolutely has a shot there.

With all of this in mind, The Postrider is proud to present its ratings for the 2020 presidential election.