It’s been about 50 days since the networks declared Joe Biden the winner of the presidential race. It will be probably a year or two before enough distance and enough research yield a trustworthy analysis of what happened. But it’s not to soon to speculate; everyone is doing it. Some preliminary conclusions and some preliminary lessons are possible.
One topic of discussion is why the 2020 polls were off. They were modestly off at the presidential level, about 3.5 points, but that’s a greater error than in 2016. And the 2020 polls were off even more in many lower races. I’ll eventually write about polling in part #2 of First Takes. Here I just address the overall results, using the polls as little as possible.
Donald Trump clearly lost, sore loser tantrums notwithstanding. Otherwise, it was about a 50:50 election between the two parties. As his last hurrah (maybe), Trump mobilized enough new and irregular voters from his base to help his party do well but not well enough for him to do well. I’m certain he would have preferred the reverse.
For Democrats, disappointed by the results beyond the White House, one lesson was that they were overconfident about mobilizing “people of color”; another is the danger of cultural overreach by big-city progressives.
What Happened?
Americans voted in huge numbers, the highest percentage of eligible voters in about a century despite efforts at voter suppression, despite fears about confrontations at polling places, and despite Covid-19. (Some, however, have argued that Covid-19 enabled greater turnout because it led many states to open up access.) Joe Biden won over 51% of the popular vote, beating Donald Trump by about 4.5 percentage points. This was not a landslide, but was a solid showing against an incumbent. The structural advantages Republicans hold in the Electoral College notwithstanding, Biden’s 4.5 margin was enough to win the College, when Clinton’s 2 point advantage over Trump in 2016 was not.
Below the presidential level, the nation split about evenly, which meant a small shift to Democrats in the Senate and a substantial shift toward Republicans in the House.
What Should We Have Expected to Happen?
Political scientists who draw on the “fundamentals” rather than the horse-race polls to predict presidential races–Is one of the candidates the incumbent? How is the economy doing? What is the sitting president’s approval rating? What are the recent trends at the state level?–would have rated the election in, say, February, as roughly a tossup (e.g., here). Trump had the incumbent advantage and he had a very good economy, but he also had unusually persistent below-50% approval ratings and a “blue wave” had hit many states two years before.
Add to these observations an understanding of voters in the 21st century. They are highly partisan and polarized compared to the prior few decades, which means that few will vote for another party’s candidates.[1] If, say, roughly 40 percent of reliable voters are dug-in Democrats and 40 percent are dug-in Republicans, then at most 20 percent are at all open to persuasion. This means that a few points one way or the other has major consequences. Additionally, many Americans are basically non- or hardly-ever voters. The persuadable and the nonvoters tend to know less and care less about public affairs than do partisan voters and basically to assume the worst of politicians.
Add also to the fundamentals the Covid-19 pandemic and the protests following the George Floyd killing. Perhaps these events should have altered expectations but how is unclear. Public perceptions of the pandemic and the protests were so polarized that they mostly mobilized partisans on each side. For committed Democrats, Covid-19 was about health; for committed Republicans, it was about jobs; the uncommitted were up for grabs. Much the same on Black Lives Matter (though sympathy for the movement ebbed as the summer wore on).
A cold calculation would lead to expecting a close race.
Why Did Biden Win?
What I mean is, Why did Biden do about 2.5 percentage points better nationally than Hillary Clinton did in 2016, enough to win five states that Trump carried in 2016? The election was close enough that many factors could be called decisive, including simply Biden being male. Essentially, Biden held on to the White, suburban, college-educated, heavily female voters who came out in great numbers against Trump in 2018. He also converted a few 2016 Trump voters and a few Republicans–few, but small percentages mattered in the swing states. Biden’s emphasis on character–that he was fighting for the “soul of the nation”–rather than on issues seems to have paid off with these swayable voters.
Thus, Biden’s major gains over Clinton came around big cities, not in the big cities. One analyst wrote, “Biden’s best hunting grounds for new voters were posh Republican counties” (see also here). Biden did not improve on Clinton’s 2016 results in the large, heavily Black center cities. In places like Philadelphia, Chicago, New York, and Detroit–Atlanta is an exception–Biden got only about as many votes as Clinton had. Meanwhile, Trump, however racist his campaign, increased his modest numbers in those same, racially diverse cities.[2]
(There is yet a simpler possibility as yet unresearched: Neither Jill Stein nor Ralph Nader was on the ballot, each of whom drew more votes than Democrats’ margin of loss in 2016 and 2000 respectively.)
Why Did Trump Do So Well?
By one standard, Trump blew it, given incumbency and the fundamentals. By another standard, he did well, given that he was remarkably unpopular and the death toll of Covid. He got the second most votes in history and the election was close.
Trump’s presidency and his campaign both played to his base. Why, many wondered, hadn’t he during his administration moved toward the center and widened his appeal? He and his team apparently believed that his base was bigger than pundits imagined, that millions of Americans who had not voted in 2016 were Trump supporters, and that registering and riling them up would counterbalance the anti-Trump mobilization that made the blue wave in 2018.
He was right. Trump brought out millions of new voters. One Democratic strategist remarked, “2018 was a wave year because our people showed up and theirs didn’t. 2020 was like a reversion to the mean because both sides showed up.” (Here, by the way, Covid worked for Trump. Republicans mounted large door-to-door campaigns; Democrats largely cancelled theirs for health safety.) In the end, this base strategy wasn’t enough to save Trump, but it was enough to save many other Republicans.
Why Did Down-Ballot Dem’s Underperform Relative to Biden and to 2018?
Presidential candidates who win a majority of the vote rarely see their party lose more than a few House seats, as just happened to Biden and the Democrats; the previous time was 1896. In part, Ronald Brownstein argues, this anomaly resulted from the growing geographical sorting by party. Democrats increasingly concentrate in overwhelming majorities in urban districts–San Francisco, for example, went 85-13 for Biden–while Republicans are distributed across exurban and rural districts where they have only mere majorities–Kern County, CA, for example, went 54-44 for Trump. Even without gerrymandering, Democrats usually need to win substantially more than half of the national vote for Congress to get only half of the resulting seats. Nonetheless, this self-sorting did not stop Democrats in 2018 from winning 40 House seats; two years later they lost about 13. Most Democratic Senate candidates also ran behind Biden (most dramatically perhaps in Maine: Biden 53%, Gideon 42%).
Some disappointed Democrats have blamed the outcome on a failed political strategy, whether it was too-progressive messaging or lousy tactics, especially in Latino areas. (More on this below.)
I think that Trump is the largest part of the explanation. There was not much ticket-splitting over all, but even a few percentage points mattered. A small handful of frequent-voting Republicans and independents voted against Trump but stuck with the GOP down ballot. The newly mobilized Trump enthusiasts voted for him but also went down the ballot to vote for other Republicans. Another part of the explanation is, indeed, messaging that painted Democrats as radicals. This was a label that Joe Biden could evade (“You know me”), but lesser-known Democrats could not. While these factors may have added only a handful of points nationwide to the GOP down-ballot candidates, that would be enough to decide many House and Senate races.
Lessons Learned
The POC Illusion
The “POC”–people of color–construction has uses. It can help build alliances against institutional racism. But 2020 suggests that, as a political tool, POC seems increasingly problematic.
One problem is that the color groups have different historical, cultural, and economic pasts and different present interests. The African-American experience is in so many ways sui generis and separable.[3] Other POCs–say, Korean-, Mexican-, Filipino-, and Jamaican-Americans–often have clashing concerns about, for example, school admissions, labor rules, and policing. College students from these groups may sometimes join together under the POC umbrella, but what excites activists at, say UCLA, hasn’t much to do with working families in Koreatown, East L.A., and South Central.
A second problem is that the POC are divided in yet another way, by generation. Decades ago, Democratic party bosses in the big cities understood that helping Irish, Italian, Polish, and Jewish immigrants move up and out of their immigrant quarters would ultimately lead many of their children to leave their parents’ Democratic loyalties behind. (Suburbanizing Jews did stay Democratic.) Assimilation was powerful then and is powerful today. The grandchildren of post-1970 immigrants, for example, typically do not speak the native language. Intermarriage becomes increasingly common. Birth rates drop toward White levels. Identification with the ethnic group fades.
Thirdly, yet another weakness in the POC strategy is the assumption that ethnic solidarity drives the politics of those who do identify ethnically. But, to adapt a comment of Thomas Pettigrew’s about racists, even ethnically-proud people have more pressing concerns than their color identity:
Faith. Latino Protestants often disagree with Latino Catholics. Jobs. In places like south Texas, many Latinos hold jobs in law enforcement and the oil industry and many worry about competition from new migrants (h/t Martin Sanchez Jankowski). Safety. In Oakland, California, Blacks were less likely to support reducing the number of police than Whites were. Ideology. A California proposition to reinstate affirmative action was supported by only a bare majority of Black voters and by only about one-third of Latino and Asian voters. (Across California counties, the key predictor of support for the affirmative action proposition is average level of education, from which you could accurately predict that I voted yes).
These non-POC issues help us understand the Latino vote. In 2020, Trump won about as large a share of the Latino vote as most recent Republican presidential candidates had gotten, substantially up from his exceptionally low performance in 2016. (Trump’s 18% of Latinos in 2016 was half of what George Bush got in 2004.) In 2020, issues like “the wall” and immigrant “invasions” were muted; the economy and the virus dominated instead. And the Trump campaign tried hard for Latino votes.
“Most Latinos identify first as working-class Americans, and Trump spoke to that,” said Josh Zaragoza, a top Democratic data specialist in Arizona. . . And then there’s the way the left spoke — or were framed by Trump’s campaign for speaking. Calls to “defund the police” . . . even using the term Latinx to describe Latinos . . . only served to puzzle many Hispanics.
Many argue that the Democrats did a lousy job reaching out to Latinos, but issues like these–as well as abortion, small-business interests, and the financial cost of Covid lock-downs–mattered more. And, Latino voters were not immune to the same anti-Black messages that the Trump campaign played to Whites.[4] In the end, Trump made his greatest gains over 2016 in heavily Hispanic areas all around the country–in Florida, Texas, California, Illinois, New Jersey–no matter the local outreach to Latinos.
For all these sorts of reasons, POC identity is not a great political motivator, at least for non-Black POC. And to the extent that ethnic identity does motivate voters, it is a game two can play. All or just about all of the House seats flipped by Republicans in 2020 were flipped by POC (or women) candidates.
As a final note, the POC strategy also counts on the immanent arrival of a “minority-majority” society when Whites will be fewer than half of the voters. It won’t be soon. In 2030, non-Hispanic Whites will still be 58% of the population over 18 years of age. Moreover, many POC are not citizens (including “Dreamers” for now). And add the fact that more and more Hispanic-Americans (and Asian-Americans) identify as sort of White. Lastly, mix in the differences in voter registration rates and you can expect the 2030 electorate to be two-thirds or more White or White-ish.
Still Waiting for the Uprising
As soon as Democrats realized that the hopes for a blue wave like that of 2018 were not going to be met–indeed, that there were going to be sizable losses in the House–they formed their usual circular firing squad. As I noted above, the moderates who lost or barely managed to win blamed the excessive demands of the progressives; progressives said that the moderates were just lousy campaigners who failed to mobilize all those potential voters yearning for dramatic change.
In arguing that there is a large but untapped pool of progressive voters, the left can point out that voters, even in purple and red states, have supported concrete progressive positions–raising the minimum wage, expanding Medicaid, legalizing pot, and so on–even while voting down Democratic candidates. Standard-issue Democrats are just not bold enough, they say.
However, this seemingly progressive voting pattern reflects an old saw in political science attributed to 1960s pollsters Lloyd Free and Hadley Cantril, that Americans are “philosophical conservatives but operational liberals.” Most Americans will support specific, concrete programs, especially once they have become familiar programs, like the minimum wage or Medicare, even if they are cut from the cloth of European-style democratic socialism. But the more the debate turns to philosophical and cultural matters–law and order, private property, taxation in general, patriotism, equal opportunity, personal rights, etc.–the more they vote conservatively. Americans, for example, favor a specific wealth tax on the very rich “to support public programs” (here and here), but only about 1 in 5 endorse the principle of “more government services if that [means] more taxes.”
The proposition that a latent majority of Americans, mostly working-class, are ready to charge to the polls if Democrats would only offer a more transformational agenda has been tested a few times recently and failed each time. One: In 2018, the Democratic candidates who captured Republican House seats were the moderates like Conor Lamb, not progressives like AOC. The latter candidates won office by defeating moderates in Democratic primaries in very blue districts. Two: In 2020, when the presidential primary contest boiled down to Biden vs. Sanders, mano-a-mano, Biden won easily. Three: 2020 was the high-turnout election the left has long been hoping for, but it was a disappointment for all but moderate Democrats such as Biden and the “5 [moderate] women who ran on their national security backgrounds to flip GOP House districts in 2018 . . . [and] won again.” (The “platoon” versus the “squad”? Also, it appears that most of the “squad” ran behind Biden in their own districts.[5]) America remains a centrist nation, at least philosophically.
Elections are not good times to shift voters’ views to yours, but a time to persuade them that you share their views. When Americans do shift significantly, it seems to be after they’ve elected a president who brings them along. Americans did not elect FDR because they favored the programs of the New Deal; Americans ended up liking those programs because they saw FDR as an empathetic president and the programs seemed to help people like them. Similarly, conservatism got more of a boost from Reagan’s presidency than vice-versa.
In 2020, the GOP raised the threat of philosophical and cultural changes that many centrist voters were not yet ready for, notably, constraining the police. They also stirred up familiar concerns about race, immigrants, government dictates about personal behavior such as masking, and presumed threats to Christianity (the Amy Coney Barrett hearings helped here).
Democrats did not repeat the down-ballot successes of 2018 probably because Trump rallied his deep base, as discussed above. But Democrats fell back even while winning the White House probably because they were associated with the most unpopular images of the left. Republicans tarred them as fellow travelers of the so-called radicals who tore down statues of Columbus and U.S. Grant, attacked property, wanted to end private health insurance, would ban gasoline immediately, and would defund the police (about which Republicans broadcast 70 different ads). This was unfair guilt by association. But fairness doesn’t matter; efficacy does.
Many political pros shared this explanation. “I don’t think there’s any doubt that wokeness, and the issues around that, helped brand the Democratic Party. The Democrats spent three months with a discourse dominated by the protests around George Floyd, racial justice and so on, culminating in the defund-and-abolish-the-police movement, which was basically of very little interest to the median voter,” said one. Obama and Biden have said much the same. (There is a reason that AOC is featured prominently in Republican ads against the two Democratic candidates in the Georgia Senate runoffs.) AOC replies, “The whole point of protesting is to make ppl uncomfortable. Activists take that discomfort w/ the status quo & advocate for concrete policy changes.” True, but meanwhile millions of centrist Americans take their discomfort out on the politicians who seem allied with the discomforters.
A fair rejoinder to this account of the Democratic setback is that the Republicans would have run the same sorts of ads whatever progressives said. Maybe, but they probably would have been less persuasive without the visibility of the progressive wing. As I said above, Biden withstood these attacks; he could get away with saying, “Come on, you know me. Do I look like a socialist?” Down-ballot Democrats had a harder time (see here and here). Even Georgia’s Stacy Abrams delicately pointed out that it was hard to sell the progressive message outside of deep blue areas.
City and Country
The surprising widening of urban-rural political–and cultural–differences in recent years got even more surprisingly wide in 2020. Metropolitan areas become more Democratic and outlying ones more Republican, irrespective of differences in race, ethnicity, age, or class. Republicans become increasingly dependent on an aging and economically stressed population of voters, while for Democrats, the message seems to be that their messaging is somewhere between unconvincing and repugnant in wide swaths of the nation.
Money
One last lesson. “Money,” as old California pol Jesse Unruh once said, “is the mother’s milk of politics.” But beyond some point–as Hillary Clinton, Jamie Harrison, Sara Gideon, Amy McGrath, and many others might now testify–a lot more political money is not any more nutritious, may just make you fat.
Update, 12/30/20
From Bruce Gyory: “The bottom line is clear: Biden won because the Never Trump Republicans gave him twice as many defections from Trump (8 percent) as Trump was able to pull away from Biden among Democrats (4 percent), enabling Biden’s clear double-digit edge among independents to carry the day.”
Update, 2/3/21
On why the Republicans did so much better than expected in the down-ballot races, especially the House, David Wasserman of the Cook Political Report points to considerations mentioned above, such as Trump drawing out rare voters who added their votes down-ballot for Republicans, but also that Republican Biden voters may have “gone home” to the GOP candidates for other offices in order to check a Biden presidency. And this:
Republicans’ attacks on “socialism” and “defund the police” were potent – and Democrats didn’t do enough to blunt them. Throughout the cycle, Democrats rolled their eyes at Republicans’ incessant ads on these themes…..[But] in a presidential cycle with plenty of voters who aren’t immersed in the policy weeds, these attacks worked. …. Relatively few Democratic challengers aired ads inoculating themselves against these attacks. Most chose to stay on “offense” on healthcare and COVID in their messaging. And, many paid the price. Says one GOP consultant, “In 2018, Dems were seen as normal. But after the rise of AOC, the primaries and ‘defund the police,’ it was easier to paint them as radical.” Democrats’ genuinely progressive challengers fared the worst ….
Update, Feb. 17, 2021
Another analysis concludes:
Despite the apocalyptic rhetoric surrounding the 2020 election, the results only deviated slightly from the 2016 contest. Voters were somewhat more partisan this past cycle; Biden improved upon Clinton’s performance, especially among independents; and the pandemic soured enough former Trump supporters for Biden to carry critical states. The close presidential result plus the down ballot Republican victories suggest that hyper-partisan elections are not going away soon.
Update, March 5, 2021
David Shor, a polling expert who had done Obama’s surveys, argues:
————————————————————————-
NOTES
[1] Here are a few of my posts on political polarization: 2012, 2017, 2018.
[2] For a discussion about Georgia turnout see Nate Cohn vs. Stacy Abrams. See also here on Trump’s gains.
[3] Nothing compares to 200-plus years of chattel slavery, a century of caste domination, and the extreme levels of racism faced by African-Americans. Native Americans and Mexican-Americans had limited experiences with slavery and peonage in America and many groups have experienced discrimination and ethnic violence (e.g., lynchings of Chinese and Italian immigrants), but the descendants of African slaves form a singular category.
[4] The General Social Survey asks respondents to explain why Blacks have done less well than Whites. From the 2000 to 2018 GSS surveys, most self-identified Hispanics rejected differences in innate ability as an explanation for the race gap, but they–like most non-Hispanic Whites–also rejected discrimination and lack of education as explanations for the gap. Hispanics were the only group of whom a majority picked “lack of will” as an explanation for Black disadvantage. (My calculations, using variables “racehisp” and “racdif1-4.”)
[5] Not all the numbers are in yet, but AOC was running only even or behind Biden in her own district [Update 2/17/21: in final results, AOC got 71.6% while Biden got 73.3% in her district]; Ihlan Omar was running considerably behind Biden in her district [16 points behind at the end]; Talib ended up slightly behind Biden; but Pressley did do 2 points better than Biden in her ring-around-Boston district. (Data for last two are from here and here, as of 12/10/20.)
MADE IN AMERICA 