This morning the New York Times released new polls of four key states, and they are all both consistent with established wisdom about the state of the race in those states, and encouraging for the Democrats:

• Arizona: Biden 49, Trump 43
• Florida: Biden 47, Trump 44
• Pennsylvania: Biden 49, Trump 43
• Wisconsin: Biden 52, Trump 41

Along similar lines, this morning the latest NBC News/Wall Street Journal national poll came out: Biden 52, Trump 42.

One interesting thing to note about this year’s polls is how few undecided votes there are. For instance, as Chuck Todd observed this morning on Meet The Press, four years ago the NBC/WSJ national poll on day -2 was Clinton 44, Trump 40.

I think this reflects several key differences between the 2016 and 2020 races. First, in 2020 third-party candidates have received little attention, whereas in that 2016 poll the Libertarian (Johnson) was polling at 6% and the Green (Stein) was at 2%. Second, an increasing proportion of the electorate is casting their vote prior to election day, reducing the potential for undecided voters. Third, evaluating whether or not to retain an incumbent is intrinsically different than deciding which of two non-incumbents to elect, leading to fewer truly undecided voters in the late days of the campaign. And last, this cycle voters aren’t needing to evaluate late-breaking news from the FBI Director casting doubts on one of the candidates.

As such, 48 hours out from the election, it seems increasingly likely that Biden will become President, in a world where the election tallies accurately reflect the voting intentions of every eligible voter who believes they voted.

I crafted that italicized phrase carefully, as there are lots of different ways in which things can go wrong in the process of tabulating election results.

For instance, as difficult as this would be to imagine had we not already lived through it, we’ve had one Presidential election whose outcome can reasonably be attributed to a “ballot design” problem, in which the ballot used in one county of an unusually close state was misleading to voters and caused them to vote for the wrong candidate. Quoting from the abstract of an article published in the American Political Science Review in 2001: “We show that the butterfly ballot used in Palm Beach County, Florida, in the 2000 presidential election caused more than 2,000 Democratic voters to vote by mistake for Reform candidate Pat Buchanan, a number larger than George W. Bush’s certified margin of victory in Florida.”

Another way in which in theory the process could go wrong is if the election tallies included the votes of ineligible voters, and those inappropriately counted votes were determinative to the outcome in one or more states. I imagine we’ll be hearing Republicans talk about “voter fraud” next week and beyond, particularly since the President has previously claimed that millions of fraudulent votes had been tabulated for his opponent in 2016, even though there seems to be little to no evidence for the concept.

And then of course there’s the theoretical possibility of interference in electoral tabulations by hackers, and the theoretical possibility of outright fraud by election officers. I am by nature a trusting soul so I’m not going to worry myself with such phantoms, but your mileage might vary, gentle reader.

No, the theme for 2020 with respect to breakdowns in the electoral process is very likely going to center around the concept of, votes that were legitimately cast under the rules in effect at the time but are not ultimately counted due to judicial action. As I had mentioned in yesterday’s post, much of this litigation activity reflects a belief by Republicans that certain actions taken by election officials to modify voting processes are inherently in violation of Article II Section 1 of the U.S. Constitution, to the extent that said actions were not enacted via state legislation.

This morning I woke up to the news that Republicans in Texas had just filed suit to invalidate over 127,000 votes that had been cast in Harris County (Houston) using “drive-thru voting” process, under the grounds that existing Texas law only authorizes the use of drive-thru voting for voters with disabilities. My understanding is that this drive-thru voting process includes the exact same voter identification procedures that would occur with normal in-person voting. As such it is difficult to make any case that this lawsuit is intended to combat “voter fraud”. Instead it represents an attempt to de-franchise people in a heavily Democratic area who believe they have already validly voted, and who may be unable to vote in person on election day even if they were to learn by then that their previous vote was to be thrown out.

Fortunately the Texas Supreme Court this afternoon rejected the request to throw out these ballots, but tomorrow morning there will be a hearing in Federal Court where the U.S. Constitutional issue becomes potentially germane. More to come, not just in Texas but surely elsewhere.

Finally, and sadly, there are increasing worries about the potential for violence in the wake of the election. Yesterday I had missed a very disturbing story out of Texas, where there is video of a so-called “Trump Train” of pickups waving Trump flags that appears to have tried to run a Biden/Harris campaign bus off the road, leading the Biden campaign to cancel three planned campaign events that day out of safety concerns. Naturally, the President approvingly tweeted a video of the incident with the phrase “I LOVE TEXAS!”, and the Texas GOP chair refers to coverage of the incident as “fake news and propaganda.” What marvelous times these are.