Election Reflection, Part II: How Did Everybody Do?

Note: this is the second of a series of Toteboard post-election reflections.

It is now exactly one month since election day, almost all of the votes have been counted (the remaining half-dozen states are scheduled to certify over the next week), and we’ve come to a place where we can look at the numbers and take stock of what went down. How did Joe Biden do? How did the democrats do? How did the pollsters do? And finally, how did the Toteboard do? Inquiring minds must surely want to know.

To address these questions, the Toteboard will employ a rating system comparable to those for reviewing restaurants, movies, and episodes of the classic Dick Van Dyke Show.

Four stars: Excellent performance, exceeding expectations.

Three stars: Solid performance, meeting most expectations.

Two stars: Adequate performance, though falling short of most expectations.

One star: Disappointing performance, falling well short of expectations.

Bullet: Complete and total failure. Shoot. Me. Now.


Three Stars.

Although falling short of the landslide many pollsters predicted, Biden nevertheless notched an impressive win. He amassed more than eighty one million votes, the first time a candidate has done so, battered Trump by more than seven million votes, took back the three big states from the erstwhile “wobbly base” (PA, MI, WI), and expanded the map with victories in two long-targeted republican strongholds (AZ, GA). Phrased slightly differently, Biden restored the so-called Blue Firewall, which was enough to put him over the top (Wisconsin was apparently this year’s tipping point), and managed to pad his victory with two extra states. Now, a cushion of two states may not seem like a whole lot, but it’s actually a fairly typical result when measured against the races from these last two decades of a closely divided electorate. Bush 43 had no cushions at all after his narrow victories in 2000 and 2004, Obama had a five-state cushion in 2008 and a three-state cushion in 2012, and Trump had a two-state cushion in 2016, so Biden is smack-dab in the center of the bell curve.

On the subject of historical parallels, the contours of Biden’s 2020 victory have much in common with those of another race in recent memory: Obama’s 2012 win over Romney. With another week of straggler votes to tally (mostly late-arriving absentee ballots, adjudicated provisional ballots, and the like), the popular vote currently stands at 51.3% to 46.9%, which is very similar to (actually slightly better than) Obama’s 51.1% to 47.2 victory over Romney. What’s more, Biden’s map looks an awful lot like Obama’s. They both won the 23 traditional blue states and DC, but then diverged on the small handful of states they added to their columns, with Obama winning Florida, Ohio, and Iowa, and Biden winning Arizona and Georgia. One could argue that Obama’s win was more impressive, as he grabbed the chunkier electoral prizes of FL and OH, but Biden earns a gold star for breaking through in states that would have seemed impossible only one or two elections ago. Regardless, as the Toteboard has mentioned before, it’s all basically a zero-sum game. If the national popular vote ratios aren’t really changing, one party’s “breakthrough” in one or two states is likely to be offset by the other party gaining ground in a different one or two states. So yeah, it’s great that Arizona and Georgia are getting bluer, but not so great that Ohio and Iowa are getting redder, which not so coincidentally makes for an almost-even electoral vote swap.

Still, the Toteboard finds it important to point out that this was hardly a resounding mandate for Biden, as his vaunted reclamation of the Blue Firewall was actually by the narrowest of margins, i.e., the biggest spread of the three was a whopping 2.8% in Michigan, which means that those three will remain key battlegrounds for the foreseeable future. What’s more, Biden couldn’t fully obscure some obvious flaws as a candidate – his sometimes foggy affect, his lackluster oratory, his shaggy-dog digressions – which may (or may not) have clipped the coattails many pundits were expecting. Make no mistake, though, Biden’s win was indeed convincing, but it’s going to take more convincing before anyone can designate either him or his heir apparent as the 2024 frontrunner.


One star.

If the Martians were to land on earth and take a look at the election results, they would think the democrats had done pretty well this time around. Apart from getting their candidate elected president, they (the democrats, not the Martians) gained one Senate seat (with two more on the bubble), won the cumulative House vote by about 3%, and came away with roughly proportionate control of the chamber despite the scourge of widespread gerrymandering. All in all, not a bad scorecard, right? Wrong. This is not what was expected from an electorate that rejected an impeached, sociopathic, morally deficient one-term president, where no fewer than seven or eight sycophantic republican Senate incumbents were thought to be vulnerable, and where democrats led the generic House ballot by better than 6% going into election day. To say that the democrats underperformed nationally is, well, an understatement. Actually, to say that it’s an understatement is something of an understatement. After winning the two Senate seats where they were heavily favored (AZ and CO), they lost every other race where they were thought to be ahead or nearly neck-and-neck (ME, NC, IA, and MT), and most of those weren’t even close (the fantasy scenarios of SC, KY, KS, and AK were all non-starters). That control of the Senate now hinges on the two Georgia runoffs is really a by-product of how poorly the democrats performed overall. And then to win back the presidency while losing 20-odd seats in the House is not only unprecedented, it’s humiliating. This may not have been Nancy Pelosi’s fault – perhaps the market just couldn’t bear two democratic waves in a row (the 2018 mid-term one was a doozy) – but as the face of the party’s House leadership, hers is the head most likely to roll. And the Toteboard wouldn’t mind seeing a new profile injected into the picture.

But the most important aspect of the dem’s tepid performance in Congressional races is that they did not sufficiently spank the repugnicans for their abuses of power, their spineless enabling of a madman, their attacks on the press, and their tribal populism. Rather than feeling chastised by this election, the bad guys are now emboldened for another cycle of obstruction, racist dog whistling, misinformation, threats to public servants, flouting of the law, and, in short, just about everything you can think of that makes politics suck. Unlike Covid, this pandemic is unlikely to be contained over the next year.


Two stars.

You’ve got to feel sorry for the pollsters – they really are the new weather forecasters. They are charged with anticipating important phenomena over which they have no control, their audiences habitually misinterpret or misapply their data, and they suffer much blame and public abuse for their perceived failings. But pollsters aren’t really in the business of predicting horseraces – their work is much more relevant for measuring changes in public opinion, shifting demographics, and so forth – and it’s not really cricket to hold them to mathematically unsound standards. But as is the case with weather forecasters, small errors don’t mean much in most cases, i.e., it doesn’t really matter whether the high on Thursday was 54 or 56 degrees, but may have tremendous implications in others. If you’re in one of those precariously situated New England towns, where the difference between 33 and 31 degrees is the difference between a soaking rain with blustery winds and an ice storm with branches collapsing on cars and houses, the accuracy of the forecast is of tremendous importance. If they say ice north of route 128, and rain to the south, and you have 20 friends on their way over for a dinner party in Lexington, you’re up the icy creek if the weather channel had a bad day. Likewise, it doesn’t really matter if a pollster misses by 3 points when the race isn’t close. But there’s a huge difference between predicting Clinton up by 2 and Trump winning by 1, as we all learned four years ago.

That said, the pollsters didn’t do such a great job this year either. They missed the popular vote by a good 3 or 4 percentage points, which while not horrible is definitely below average, but the bigger problem was that they missed several key states for the second time in a row, by similar significant margins, in favor of the same party. And so it’s a pretty big deal that they underestimated the repugnican vote in Michigan by 4.5% in 2016 and 5.2% in 2020, in Wisconsin by 6.3% in 2016 and 7.7% in 2020, in Pennsylvania by 3.9% in 2016 and 3.5% in 2020, in Iowa by 6.7% in both 2016 and 2020, in Ohio by 6.7% in 2016 and 7.5% in 2020, and in North Carolina by 4.5% in 2016 and 3.0% in 2020. And while they only missed Florida by 2.0% in 2016, that ballooned up to 5.8% in 2020, in large part from having misread the Hispanic vote, and so they have a lot of splainin’ to do on that one. On election eve, several analysts noted that Biden's polls were such that he could survive a 2016-style polling error, and he did, but not without some very tense moments. And this isn’t even including the equally dramatic bungles in the Senate and House.

And yet, in spite of all the not so little things (like the decline of landlines and the rise of a global pandemic) that made their job a real crapshoot this year, the pollsters did have their bright spots, both individually and collectively. For example, everyone wondered if the late-breaking Seltzer Poll in Iowa was a weird outlier, but it turned out they had a good ear for the cornfields after all. Likewise with the ABC News Washington Post poll in Florida. More interestingly, the cumulative polling averages predicted with deadly accuracy Georgia’s ever-so-slight Biden win. In fact, in might be helpful to consider that while the pollsters pretty much blew the numbers, they nevertheless got the narrative right. That is, they predicted that Biden’s best shot was to rebuild the Blue Firewall (correct on that) and that the candidates would be slugging it out in Arizona, Florida, Georgia, and North Carolina (correct again). So even while the pollsters are licking their wounds for bad percentages, their work made it possible for Larry Sabato to correctly predict 49 out of 50 states, including Arizona and Georgia for Biden, and Florida for Trump (his one clinker was North Carolina), and for Nate Silver to correctly predict 48 of them (missing NC and FL). And of course, if it weren’t for the polls, the Toteboard wouldn’t be able to provide the kind of analysis it does.


Because the Toteboard does not really predict election outcomes as much as provide frameworks for understanding elections – it only provides horse-race odds for primary outcomes – the Toteboard will forgo rating its own performance. But it will examine how useful its frameworks were, and consider ways to noodle with them for future elections. Previously, the Toteboard employed its variations on what could best be called the “Color Code,” i.e., sorting into red, blue, purple, indigo, and burgundy states, while expanding the lexicon to include what are now household terms (for Toteboard readers, anyway) like “wobbly base” and “aspirational purple.” But for this weird election coming at the end of Trump’s weird presidency, the Toteboard adopted a pair of other models as well, i.e., the “Regional Cluster” model, and the “Triad” model, in part for tracking interesting changes in the electorate, in part for helping to navigate election night itself. As so, we’ll take separate looks at all three of these frameworks.

The Color Code: This model identified the Blue Base as 17 states (CA, CT, DE, HI, IL, MA, MD, ME, MN, NH, NJ, NM, NY, OR, RI, VT, WA) plus DC, totaling 205 electoral votes, all of which Biden carried handily. He also carried the 28 EV’s of the Indigo Three (CO, NV, VA), marking the fourth presidential election in a row where those states voted democratic (though Nevada was admittedly a little closer than we might have liked). Thanks to this recent electoral history, and the fact that all three states boast democratic governors and two democratic senators, the Toteboard is now comfortable inviting those three to join their 18 friends in the Blue Base, bringing the total there to 233 EV’s. Congratulations, Colorado, Nevada, and Virginia! This would be a good time to celebrate with some Rhode Island calamari! Biden also won three out of the four states in the Wobbly Base (PA, MI, and WI, but not IA), though those victories were so tenuous that the Toteboard feels obligated to rebrand Pennsylvania, Michigan, and Wisconsin as the new Indigo Three (sorry about that, guys), i.e., blue-leaning purples that will be heavily contested. It is also reassigning Iowa as Burgundy (joining FL, NC, and OH, all of which Biden lost), though to be honest, both Iowa and Ohio have been so bad for democrats lately they may soon lose their Purple status. Still, the Toteboard prefers not to move in haste after only one or two election cycles, which leaves us wondering exactly what to do with the Aspirational Purples of Arizona and Georgia, which now no longer seem merely aspirational. On the one hand, one election doth not an Indigo state make, but on the other hand, the trendline seems to be so good in both states, with Arizona going for Biden and electing two democratic senators, and Georgia on the edge of delivering the Senate to the democrats and having Stacey Abrams in the wings for another gubernatorial run, that it’s tempting to bypass Burgundy and assign them straight to Indigo. For now, the Toteboard is going to hedge its bets and treat them both as generic Purple states (or the New Purples), with the understanding that both could blue as quickly as Virginia . . . or revert to red form after a blue anomaly or two (like North Carolina). For the record, the mathematically fetishistic Toteboard should acknowledge that tracking the states as they change colors like liquids in a chemistry experiment has actually been one of the most intriguing aspects of election-watching.

The Regional Clusters: To quote the 9/3/20 Toteboard: “A generation or so ago, the country was more or less divided along some fairly clear lines: the republicans had a lock on the south and Plains states, while the democrats had the northeast and the west coast, and most of the contested territory was in the blue-tinged upper-midwest. But now, there are at least three “hot spots” on the map, i.e., clusters of states that are emerging (or pre-emergent) battlegrounds. Yes, there’s the old Rust Belt Quadrumvirate of Pennsylvania, Michigan, Ohio, and Wisconsin, but there’s also what might be called the Peripheral South (as opposed to Deep South) states of Florida, Georgia, and North Carolina, and the Southwestern Cactus League of Arizona, New Mexico (which is technically blue, but Trump is contesting it), and Texas (which is red, but may not remain that way).” In fact, all three of these areas produced interesting, mixed results. Biden won most of the Rust Belt Quadrumvirate, but only narrowly; he held New Mexico and picked off Arizona in the Cactus League, but made only nominal inroads into Texas; he yanked Georgia out of its southern miasma and gained some (but not enough) ground in North Carolina, but underperformed Clinton in Florida. In short, it looks like all three regions will remain hot-spots for elections to come, where the general trend seems to be democrats slowly losing ground in the Rust Belt, gaining ground in the Cactus League, and probably breaking even after slight realignment in the Cosmopolitan South (note the new and improved nomenclature). The late religious studies scholar Jonathan Z. Smith had memorably written that “the historian’s task is to complicate, not to clarify.” And indeed, the electoral map is growing more complicated, and less clear. It’s a good thing we’ve all got the Toteboard.

The Triad Model: As a tool devised only for the purpose of navigating election night in the light of the most recent polling (and covid), the three Triads – the Firewall Triad (PA, MI, WI), the Insurance Triad (AZ, FL, NC), and the Landslide Triad (GA, IA, OH) – will be retired for now, perhaps to be recalled and revised as needed in the future. This grouping more or less did its job, though the uneven vote compilation times made that a little less obvious. That is, the best case scenario was that Biden would carry the Firewall Triad, and we could all wind down with an episode of “Kim’s Convenience” and turn in early. And that was almost the case, as Michigan and Wisconsin did get called on election night, but Pennsylvania took longer, a lot longer, and so we had to dip into the next tier for reassurance. Unfortunately, the Insurance Triad didn’t start out particularly promising, but Fox News’s unexpected call on Arizona was big, effectively announcing that Biden had won the election whether or not he came back in Pennsylvania (as long as Nevada behaved itself) . . . . except that the other networks didn’t follow suit, and so we could only be sort of tepidly confident that it was over. Finally, by the time the needle turned the other way in Georgia, it may not have indicated a landslide, but it did reinforce the emerging narrative that Biden had won and that it would take a lot of legal shenanigans in a lot of states for Trump to get this one overturned.

Of course, the election isn’t really over, and Georgia will be getting more attention over the next month than it got during the 1996 Olympics. Indeed, one can feel some of the electricity in the air, as the bright signs along DeKalb Avenue inspire the coalition to “VOTE ONE MORE TIME!” And just in time for all the excitement, the next Election Reflection will include the Toteboard’s status report on the Georgia Senate races. Until then, stay tuned!

PS: The Toteboard would like to hear from you! Here is your chance to get into the prognostication game. Who do YOU think will be the democratic and repugnican nominees for president in 2024? Of course, the Toteboard has its own ideas, but it might be fun to hear what everyone else is thinking. Send your predictions to the Toteboard sometime between now and Biden’s inauguration, and we’ll announce the results shortly thereafter. And by the way, whoever gets it correct will win the prize of Lifetime Gratification.