Now that I’ve had a chance to watch at least one network’s (Fox News’) coverage of the 2020 election, one thing that strikes me is how quickly Donald Trump lost his popular vote lead in the last election.

It was in 2004 that networks began displaying the national popular vote with some regularity over election night. In 2020, it seems that networks gave much less importance to the national popular vote. Fox News showed it sparingly, and without percentage figures until after midnight on the East coast. (Lest one think that is because of partisanship, it doesn’t appear as though NBC showed it at all, although I’m not sure if they did so in 2016. …


Iowa: the new anti-bellwether (for now)

In 2011, , given that it was the only state to have voted for losers of different parties within the previous six elections (Dukakis in 1988 and McCain in 2008). It turned out that West Virginia had simply transitioned from being solidly blue to being solidly red. Henceforth, at least for the foreseeable future, West Virginia is likely to ‘always [be] an outlier on one side of the partisan divide’ (which was Sullivan’s quibble with Josh Goodman’s contention that Utah was the best anti-bellwether). …


In many ways, Donald Trump put in an impressive performance in the 2020 election. However, if he was going to win re-election, it was likely going to resemble 2004, when George W. Bush made those parts of small-town and rural America that had already long been red, even redder; made those parts of small-town and rural America that had been blue until he flipped them in 2000, still redder; and made still further inroads into then-traditionally blue parts of small-town and rural America.

To some degree, Trump did do this. But he didn’t do so to the same extent as Bush did in 2004 (or as Bush did in 2000, or as he himself did in 2016). In some places, he even went backwards, as Biden narrowly took back some of the counties that had been critical to Trump’s win in 2016 and narrowed his margin in others. In general, he didn’t carry those that he did carry by the kinds of margins that Obama in 2012, Kerry in 2004, or Gore in 2000 did. But in carrying them at all — and in largely denying Trump any further inroads — Biden’s effort resembled Romney’s in 2012. In 2012, Romney did manage to whittle away at Obama’s margins in most of the large suburban counties he had won in 2008, even managing to narrowly win back a few (such as Chester, PA, Douglas and Lancaster, NE, and Marion, OR). Meanwhile, as the trend of working class white America towards the GOP continued in 2012 (probably not as a function of Romney’s candidacy), so too did the trend of the educated, affluent suburbs towards the Democracy in 2020 (again, probably not as a function of Biden’s candidacy). …


‘Large state’ is a term that needs some precision. It might not seem all that important, but in 2008, ‘Ohio’s the only large state outside the South that [Bush] carried.’ Insofar as Will is making a point about national politics, it is actually important to have a non-fuzzy definition of a ‘large state’.

Now, a number of analysts’ intuitions appear to have converged at around 15 electoral votes or so as the threshold for a ‘large state’. In 1988, Walter Manley II analysed Dukakis’ and Bush Sr’s chances in the seven contested (explicitly justifying leaving out New York and Florida on the basis of their safely Democratic and Republican statuses, respectively). The seven states he analysed, together with New York and Florida, were the nine biggest electoral prizes that year, ranging from 16-EV New Jersey to 47-EV California. (The next-largest electoral prizes after New Jersey in 1988 were Massachusetts and North Carolina, each with 13 electoral votes at the time.) In 2008, Rhodes Cook defined a ‘megastate’ as a state . …


Almost every state — and all the close states — have . Which means we can figure out what 2020’s ‘magic number’ is.

An election’s ‘magic number’ is . For example, the ‘magic number’ in 2000 was 269 voters in one state, Florida. If that number of voters in Florida had switched from Bush to Gore, Gore would have won Florida by one vote — 2,912,522 to 2,912,521 — and therefore the election.

In 2020, a switch of 40,831 voters in Pennsylvania, 6,336 voters in Georgia, and 5,229 voters in Arizona would have re-elected Trump. That’s a total of 52,396 voters in three states. …


Shortly after the 2012 election, defined Missouri, Illinois, Indiana, and Ohio as the ‘Bellwether Belt’, because these four states’ collective behaviour had tended to predict election outcomes over a historically long term. However, there are a number of possible groups of states whose collective behaviour has predicted election outcomes, so I will be using ‘bellwether belt’, in a general sense, to refer to any such group of states.

Perhaps the oldest ‘bellwether belt’ is the ‘Route 30 states’. Route 30 runs from the Atlantic to the Pacific, but ‘the Route 30 states’ seem to have generally been defined as the first five states along Route 30 (excepting the tiny section in West Virginia) starting from the Atlantic coast: New Jersey, Pennsylvania, Ohio, Indiana, and Illinois. For example, on Halloween 1976,


Votes are still trickling in from the 2020 election, but it appears that there is enough certainty to say that . It last backed a loser in 1976, when it voted for Gerald Ford.

What’s interesting is that there were a number of bellwether counties, and they were in voting to re-elect Trump. Perhaps the most famous is Vigo County, Indiana. But if a bellwether county is one that has voted for the winner in at least ten presidential elections in a row, there were 19 bellwether counties as of Halloween of this year. (This is the definition used by Dave Leip in his 2001 essay on ; it is also the definition used by Claire Garofalo in her . Ten elections covers a period of 36 years, which has some significance in , although I’m not sure that’s how the definition was derived.) …


, I discussed what state was most analogous to Ohio for the Democracy, in the sense that it was nearly the case (or had remained the case for the longest) that no Democrat had won without it.

But is the fact of no Republican’s having won without Ohio itself all that interesting or significant? Is Ohio somehow representative of the Republican Party’s historic base?

Well, it is true that no Republican has won without Ohio, and it is the only state whereof this is true, although there were other states whereof this was true until fairly recently. Until 2004, no Republican had won without New Hampshire; and until 2000, no Republican had won without Vermont or without Illinois. Vermont and New Hampshire are fairly small states (although West Virginia, , is also fairly small). Illinois is about the same size as Ohio, and in 2000, that no Republican had won without Illinois. However, even then, while it remained true that no Republican had won without either Ohio or Illinois, there seemed to be more general interest in the fact that no Republican had won without Ohio (with Woodruff’s comment being relatively isolated). …


Every election, we can rely on someone remarking that no Republican has ever won the White House without Ohio (although, if Ohio is drifting toward becoming a solid red state, that may become less pertinent). These remarks carry the somewhat superstitious implication that Republicans can’t win without Ohio. (After Ohio was called for George W. Bush in 2000, Jeff Greenfield remarked that Ohio was a state Bush .)

Presumably, however, the significance of no Republican having won without Ohio lies, at least implicitly, in Ohio’s being particularly representative of the Republican Party’s historic base: if it is losing Ohio, it is not merely losing Ohio’s electoral votes, but also losing correlated voters in other states, voters who have formed an ancestral part of the Republican base. …


In 2011, ‘Most Republican State’ to refer to the state that had given the Republican presidential nominee his highest vote share after a given election:

While the title of Most Republican State was migrating from Vermont to Utah, it stopped in this Farm Belt state [Nebraska], which was Richard Nixon’s best in 1960 (62.1%) and in 1968 (59.8%).

The peregrinations of the titles of ‘Most Republican’ and ‘Most Democratic State’ are quite telling in terms of the changes in the parties’ coalitions.

The location of the title of ‘Most Republican State’ had a preternatural stability from 1856 through 1916, and to a lesser extent from 1856 through 1956. In every election from 1856 through 1916, with two exceptions, Vermont was the Most Republican State. One exception was the wartime election of 1864, when Kansas, the Jayhawker State, was Lincoln’s best state. The other, 1912, was less of an exception than it first seems. While Utah gave Taft his highest vote share, it was still a quite low 37.42% — low enough that, while Charles Evans Hughes improved on it slightly in 1916, he still lost the state by over 20%, failing to carry a single county. This suggests that Taft’s relatively high vote share in Utah may have been the function of greater partisan loyalty on the part of those of its residents who were Republican, even if they were a minority. …

Matt Lennox

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store