Thursday, February 27, 2020

An Outbreak of Common Sense on Electoral Strategy?

Ron Brownstein has an excellent article on CNN on the fallacies of a turnout-based strategy, such as Sanders has repeatedly advocated, for winning the 2020 election. He effectively summarizes the key points I and others have made against this strategy. But perhaps the most interesting part of the article is a series of quotes from Sean McElwee going on record that he, too, thinks the turnout strategy is bats.
This is a bit surprising since McElwee rose to prominence through his advocacy of the "abolish ICE" slogan and his fervent support for the AOC-brand of strenuous progressivism. He went on to co-found Data for Progress, whose work has generally seemed aimed in the same direction. That said, they have done some good work and are to be commended for fighting their battles on progressive strategy with data instead of dogma and assumptions.
Perhaps it was the experience of staring over and over at the actual data that has led McElwee to part company with orthodox Sandersism on this issue:
"[James] Carville has emerged as a leader among Democrats concerned that nominating Sanders will doom the party to defeat against Trump and put the House majority at grave risk as well. Unlike Carville, Sean McElwee, founder of the liberal-leaning group Data for Progress, believes Sanders can find a pathway to victory against Trump by attracting working-class voters across racial lines. But McElwee agrees with Carville that no candidate, Sanders included, can bet on winning mostly by transforming the nature of who votes.
"I think that all campaigns are incentivized to portray themselves as doing something unique and groundbreaking and really changing the structure of turnout." McElwee says. "But turnout is a pretty durable attribute and it tends to correlate with intrinsic human identities: Older people tend to vote at much higher rates than younger; college educated vote more than non-; homeowners vote more than renters. It is really, really hard using the tools available to campaigns to change that."
This dispute has profound implications as Democrats' assess Sanders' potential viability as a general election candidate. The Democratic front-runner brushes off concerns about whether his agenda will alienate swing voters by insisting he can compensate by bringing in millions of new voters to overwhelm them."...
If Sanders can't win a general election by changing the electorate, as these Democratic experts believe, that means he, like any other potential nominee, would need to win primarily by converting swing voters. Though Sanders always stresses mobilization, especially of young people, some of his supporters -- and advisers -- believe that he would be more likely to beat Trump by attracting working-class voters across racial lines, including whites, African Americans and Hispanics.
"If you are hitching your wagon on a youth quake [of new voters] you are in a bad place," says McElwee. "But Bernie doesn't have to hitch his argument on that. Bernie has a persuasion argument for swing voters."
Now I have my doubts about Sanders' ability to appeal to swing voters--or even interest in doing so--but at least we're aiming at the right target here! Sanders could indeed be the nominee and he could indeed win, but to do so he will have take some of this wisdom on board.
He will also have to deal with these problems, as summarized by Brownstein:
"* [S]ubstantial resistance to his unprecedented tax-and-spending plans among the college-educated suburbanites who moved toward the party in 2018 because of their distaste for Trump. (A recent analysis using 170,000 interviews from the nonpartisan Nationscape survey found that Joe Biden and Sanders posted similar leads over Trump overall in tests of 2020 sentiment, but that the former vice president ran much better among college-educated white voters.)
* [R]esistance to many of his views on issues relating to race and culture. Polls last year by the Marist Institute found that most noncollege whites supported such core Sanders economic proposals as a wealth tax on large fortunes and raising the minimum wage. But they registered overwhelming opposition to other ideas he's embraced: In one Marist survey, 67% of noncollege whites opposed eliminating the death penalty, 72% opposed decriminalizing illegal border crossing and 76% rejected providing subsidized health care to undocumented immigrants. In the Marist polling, a majority of noncollege whites have also consistently opposed one of Sanders' core policy proposals: a single-payer national health care system that would eliminate private insurance with only a very few exceptions."
It's a steep hill to climb once you discard the turnout mythology. Perhaps it can be done, but it will require Sanders and his advisors to stop getting high on their own supply.
Bernie Sanders, the clear front-runner for the Democratic presidential nomination after his commanding victory in Nevada on Saturday, often says the principal reason he can beat President Donald Trump in a general election is that he will massively increase voter turnout.

Tuesday, February 25, 2020

The Turnout Myth, Part 4

Well, as Ronald Reagan said in another context and another debate long ago: "There you go again". In the debate tonight, Sanders once again gave his electability pitch and it was exactly the same as it has been all campaign: a gigantic, humongous surge of turnout that will sweep all the Democrats' troubles away. I've been whacking away at this ridiculous assertion as best I can and, coincidentally released the same day as the debate, we have a new and thorough political science study (summarized on Vox) that makes clear just how heroic and unrealistic Sanders' assertions are.
The study, by David Broockman and Joshua Kalla of Berkeley and Yale respectively, essentially shows that Sanders' seeming electability in trial heat surveys--where he runs as good or better than "moderate" candidates--is attributable to two dynamics: (1) Sanders losing voters to Trump relative to moderate candidates but (2) making this up as people who say they are neither/third party/wouldn't vote with a moderate candidate move to Sanders. The net of these two trends keeps Sanders afloat and "electable".
Could this happen in real world? Probably not. The problem is that the implied turnout increase for young voters--who are the ones who come off the sidelines in a Trump-Sanders trial heat-- is wildly implausible. As the study notes, Sanders would have to generate a larger increase in turnout among young voters than Obama managed to generate among black voters in his historic 2008 election. Looked at another way, youth turnout would have to not just go up 11 points but 11 points more than everyone else goes up in the 2020 election. So if turnout goes up 11 points in the rest of the population in 2020, youth turnout would have to go up 22 points. you get the idea.
So, more evidence that Sanders' theory of the case on how he would beat Trump is furshlugginer (look it up).
Some excerpts from the researchers' Vox summary of the study:
We found that nominating Sanders would drive many Americans who would otherwise vote for a moderate Democrat to vote for Trump, especially otherwise Trump-skeptical Republicans.
Republicans are more likely to say they would vote for Trump if Sanders is nominated: Approximately 2 percent of Republicans choose Trump over Sanders but desert Trump when we pit him against a more moderate Democrat like Buttigieg, Biden, or Bloomberg.
Democrats and independents are also slightly more likely to say they would vote for Trump if Sanders is nominated. Swing voters may be rare — but their choices between candidates often determine elections, and many appear to favor Trump over Sanders but not over other Democrats.
Despite losing these voters to Trump, Sanders appears in our survey data to be similarly electable to the moderates, at least at first blush. Why? Mainly because 11 percent of left-leaning young people say they are undecided, would support a third-party candidate, or, most often, just would not vote if a moderate were nominated — but say they would turn out and vote for Sanders if he were nominated....
[T]he “Bernie or bust” phenomenon appears almost entirely limited to left-leaning young people, who are usually a small share of the overall electorate. This stands in contrast to many theories of Sanders’s electoral appeal: For example, whites without a college degree — a demographic some speculate Sanders could win over — are actually more likely to say they will vote for Trump against Sanders than against the other Democrats. The same is true of the rest of the electorate, except left-leaning young people....
The case that Bernie Sanders is just as electable as the more moderate candidates thus appears to rest on a leap of faith: that youth voter turnout would surge in the general election by double digits if and only if Bernie Sanders is nominated, compensating for the voters his nomination pushes to Trump among the rest of the electorate.
There are reasons to doubt a Sanders-driven youth turnout surge of this size would materialize. First, people who promise in surveys they will vote often don’t, meaning the turnout estimates that Sanders’s electability case rests upon are probably extremely inaccurate. Second, such a turnout surge is large in comparison to other effects on turnout. For example, Sanders would need to stimulate a youth turnout boost much larger than the turnout boost Barack Obama’s presence on the ballot stimulated among black voters in 2008....
And this enormous 11 percentage point turnout boost is only enough to make Sanders as electable as the more moderate candidates, given the other votes he loses to Trump. For him to be the most electable Democratic candidate based on his ability to inspire youth turnout, Sanders’s nomination would need to increase youth turnout by even more....
There is no way to be sure whether Sanders’s nomination would produce this historic youth turnout surge — but it seems doubtful. Turnout in the 2020 primaries so far has not exceeded 2008 levels, including among young voters. If anything, research suggests the opposite is more likely to occur: In response to an extreme Democratic nominee, Republicans could be inspired to turn out at higher rates to oppose him.
New research suggests Sanders would drive swing voters to Trump — and need a youth turnout miracle to compensate.

Monday, February 24, 2020

Can Sanders Beat Trump?

It is certainly possible. But that's really not the right question. The right question is: how likely is it that Sanders would beat Trump if he were the nominee?
Jon Chait makes this point with admirable clarity in his latest column:
“The truth is we are all clueless about what voters want or will accept,” argues conventional-wisdom-monger Jim VandeHei, in a signal of how deeply the anti-probabilistic fallacy has spread. It is true that there is uncertainty attached to every outcome. The talking heads who guarantee Sanders will lose are wrong — any nominee might win, and in a polarized electorate, both parties have a floor of support that gives even the most toxic candidate a fighting chance. In 2016, Trump was the most unpopular candidate in the history of polling, but he squeaked into office because everything broke just right for him. It could happen for Bernie, too.
But to concede that we cannot be certain about the future does not mean we know nothing. An imperfect comparison might be to predicting the outcome of sporting events. You don’t know the outcome in advance, but it is usually possible to make probabilistic predictions. Those predictions are wrong all the time. But it would be silly to conclude that, just because upsets happen, every game should be treated as a coin flip. A huge amount of pro-Sanders commentary is based on simplistically conflating the correct claim that we lack perfect clarity with the incorrect claim that we have no clarity at all."
With that in mind, what do we know that might shed light on how Sanders would do against Trump? First, of course, there is the trial heat polling. That polling, according to RCP averages, has Sanders and Biden running ahead of Trump nationally by essentially identical amounts and both ahead of other tested Democratic candidates.
The same pattern with Biden and Sanders relative to the other Democratic candidates can be seen in swing state polling, with the difference that Biden generally generally runs a little bit better than Sanders in most swing states. You can see this both in the RCP trial heat averages and in preliminary state-level breakdowns of the Voter Study Group Nationscape survey (more than 170,000 interviews so far, 6000 nespondents per week)
This suggests that both Sanders and Biden, neither of whom has name recognition problems, are currently capturing anti-Trump, pro-Democratic preferences fairly efficiently. Put another way, simply hearing their names and knowing who they are, does not, at this point, deter large numbers of respondents from expressing pro-Democratic sentiments.
But in a general election campaign, of course, the Trump campaign will be working strenuously to sow doubts about the Democratic candidate and convince undecided voters and those with soft Democratic preferences that Trump, whatever objections such voters may have to him, is by far the lesser evil when compared to the Democrat. This is where Sanders will run into trouble, since since he is poorly set up to parry such attacks among persuadable voters.
David Leonhardt summarizes his problem succinctly:
"[Sanders] has taken a nearly maximalist liberal position on every major issue. It’s especially striking from him, because he has shown over his career that he grasps the importance of building a coalition.
Sanders once won over blue-collar Vermonters with help from a moderate position on guns. “We need a sensible debate about gun control which overcomes the cultural divide that exists in this country,” he said in 2015, “and I think I can play an important role in this.” He was also once an heir to organized labor’s skepticism of large-scale immigration. “At a time when the middle class is shrinking, the last thing we need is to bring over in a period of years, millions of people into this country who are prepared to lower wages for American workers,” he said in 2007.
Now, though, Sanders has evidently decided that progressives will no longer accept impurities — or even much tactical vagueness. He, along with Elizabeth Warren, has embraced policies that are popular on the left and nowhere else: a ban on fracking; the decriminalization of border crossings; the provision of federal health benefits to undocumented immigrants; the elimination of private health insurance.
For many progressives, each of these issues has become a moral litmus test. Any restriction of immigration is considered a denial of human rights. Any compromise on guns or health care is an acceptance of preventable deaths.
And I understand the progressive arguments on these issues. But turning every compromise into an existential moral failing is not a smart way to practice politics. It comforts the persuaded while alienating the persuadable.
F.D.R. and Reagan understood this, as did Abraham Lincoln and many great social reformers, including Frederick Douglass, Jane Addams, Martin Luther King Jr. and Cesar Chavez. Strong political movements can accept impurity on individual issues in the service of a larger goal: winning."
That's the nub of his problem right there. He really is extremely vulnerable to brutal attacks from his Republican opponent, which will require unusual deftness and savvy to counter successfully. So far, we haven't seen a Sanders who seems capable of doing that.
Of course,Sanders does have a response to the potential difficulty summarized here: turnout, turnout, turnout! But as I and others have shown, this is a chimera. If Sanders is to beat Trump, he'll have to it the old-fashioned way: convincing many voters who don't adore him that he is indeed a superior choice when compared to Trump.
Who are these voters? Some clues may be found a recent piece by Patrick Ruffini based on Nationscape data. Ruffini finds that while both Biden and Sanders have solid leads over Trump in the national data, their coalitions are not identical. Specifically, Sanders does quite a bit better than Biden among young voters but lags seriously lags behind among voters over 45. And while Sanders is comparably strong among nonwhite voters and lags Biden only slightly among white noncollege voters, he trails Biden's performance by 8 points among white college voters.
If Sanders is the nominee and wants to maximize his probability of beating Trump, he is going to have to face up to these difficulties. If not, I fear we're in for a long and painful next four years.
About this website

Nevada: Bernie 2020 Vs. Bernie 2016

It's worth stressing that while Sanders is in the catbird seat for getting the nomination it is not because he is drawing overwhelming support from voters. Though he is doing very well, the most salient reason why he is in such a dominant position is the incredibly fragmented nature of his opposition.
The Nevada numbers are striking in this regard. First, take the overall vote numbers. In 2016, Sanders, judging from the 2016 entrance poll, had around 46 percent of the first preference vote and wound up with 47 percent of the county convention delegates. This time around he had around 33 percent of the first preference vote....and got 46 percent of the county convention delegates.
Looking at the entrance poll breakdowns in 2016 and 2020, Sanders did indeed do very well among Latinos this year, receiving 51 percent support. But in 2016, he received 53 percent, so Sanders held his 2016 support among this group--impressive in a multi-candidate field--but he did not increase it. Similarly, Sanders received 27 percent black support this year, an increase, but not a large one, over his 22 percent support in 2016.
As these figures imply, Sanders did considerably less well this time around among white voters. In 2016, he was supported by 49 percent of whites; this time he drew only 29 percent support. Breaking this down by education, he fell from 46 percent in 2016 to 24 percent among white college graduates and from 52 percent to 35 percent among noncollege whites.
In light of the Nevada results, it certainly seems like the non-Sanders elements of the party could prevail against him if they unified against him. But that does not seem likely to happen anytime soon; the next contest is in South Carolina and no one seems in any hurry to drop out. Indeed the contest will include the ludicrous candidacy of Tom Steyer, who could draw as much as 18 percent of the black vote. Then it's Super Tuesday with the other billionaire doofus, Mike Bloomberg, in the mix to further fracture the non-Sanders vote.
Now, if you are bullish on Sanders' chances to beat Trump, this is a perfectly fine development. But if you are very doubtful of his chances. this could be a disaster in the making. In the latter case, you are at the mercy of the wisdom of the various non-Sanders candidates and their ability to clearly assess the political situation.

Friday, February 21, 2020

The Michigan-Pennsylvania-Wisconsin Story, According to Quinnipiac

This week, Quinnipiac released polls of the three key swing states of Michigan, Pennsylvania and Wisconsin--the three states that put Donald Trump in the White House.
So: how'd the Democrats fare in the trial heats? In two the three states, Michigan and Pennsylvania, tested Democrats all beat Trump by varying amounts. In Wisconsin, however, all tested candidates lose to Trump by varying amounts. This is consistent with the general patterns we've seen in other states polls, with Wisconsin being the weakest of these three states for the Democrats.
Looking at the crosstabs Quinnipiac provides for all three polls is instructive, especially the white college and white noncollege tabs. It's highly likely that trends among these two groups will determine the general election outcomes.
For simplicity's sake, I focus on Biden's tabs in the three states (Sanders and Biden perform very similarly against Trump in Michigan and Wisconsin, while Biden has an advantage in Pennsylvania.
Using the States of Change data I can compare the margins in the Q-polls to those among these two groups in 2016. The results are quite interesting.
First, the good news. In Pennsylvania Biden is running ahead of Clinton's performance among both white college (+25 vs. +9 in 2016) and white noncollege voters (-16 vs. -29). In Michigan, the big improvement is among white college voters (+16 vs. -2); the figures for white noncollege are virtually identical (-22 vs. -21).
But in Wisconsin, the bottom drops out. While white college performance for Biden is fine (+14 vs. +15 in 2016), white noncollege shows a sharp drop (-26 vs. -19).
One doesn't have to 100 percent believe the Q-poll figures to see that they provide considerable food for thought.

Thursday, February 20, 2020

Is Trump's Approval Rating Really Going Up?

I've written about this before but there are reasons to be skeptical that his remarkably stable approval ratings are suddenly headed upwards. The excellent G. Elliott Morris explains in The Economist.
"[E]ven after correcting for demographic biases, pollsters’ data can still be unrepresentative. They may have the right shares of Latino voters and boomers, but nevertheless have too many Republicans or Democrats. This concern is pronounced when an event causes especially good, or bad, news for a political party. At such times surveys can suddenly be swamped with partisans who are eager to voice their love, or hate, for the president.
In the wake of Mr Trump’s acquittal in the Senate, pollsters suspect that such a bias could be affecting polls. Courtney Kennedy, the director for survey research at the Pew Research Centre, says that there is a “strong possibility” that the recent uptick in Mr Trump’s ratings has a wave of optimistic Republicans as its source. She says that outlets can control this problem by adjusting their data to have the correct shares of Democratic- and Republican-leaning voters, but the idea is relatively new and few pollsters have data good enough to perform such corrections.
The Economist’s analysis of polls taken during Mr Trump’s impeachment proceedings affirms Ms Kennedy’s suspicion. In polls that weight their data to represent America’s partisan balance or the results of the 2016 election, the share of adults who approve of Mr Trump’s job as president has risen by half a percentage point since impeachment proceedings began in earnest last October. But in polls that do not, Mr Trump’s ratings have increased by over three percentage points."
Now if we only had a candidate who could take advantage of the fact that Trump is still really, really unpopular and likely to remain so....

Wednesday, February 19, 2020

What If Everyone Voted? Be Careful What You Wish For!

I've questioned the wonder-working powers of high voter turnout before but, shockingly, not everyone has agreed. But here is some more evidence undercutting the more-turnout-will-solve-everything thesis from a massive study by the Knight Foundation. There's a great deal in their report, including a very interesting typology of nonvoters, both their characteristics and reasons for not voting, which suggest a complex phenomenon not reducible to voter suppression and/or insufficiently radical candidates.
But, regardless of motivation,what if all those nonvoters really did vote? Surely the Democrats would kick Trump's ass back to Mar-a-Lago for good. Sorry, it's not that simple, as the chart of key swing states below shows. Democrats would benefit some in the national popular vote but wouldn't be helped sufficiently in the Electoral College to take Trump out.
If this doesn't make you question the turnout mythology currently popular in Sanders wing of the Democratic party, I don't know what will.

Sunday, February 16, 2020

Could Bernie Bros (and Gals) Cost Democrats the Election If Bernie Doesn't Get the Nomination?

The assumption here seems to be that Bernie voters will be so ticked off if he doesn't get the nomination that they'll defect in huge numbers voting third party, voting for Trump or just not voting at all. Result: Democratic defeat. It's kind of like that old National Lampoon cover: "Buy this magazine or we'll shoot this dog".
I think people can calm down about this one. If Democratic voters refuse to "buy" Bernie, I think it's quite unlikely there will be mass defections. After all, in 2016, the defection rate of Bernie supporters--who you may remember will apoplectic by the end of the process--was actually relatively low: according to CCES data, 17 percent of Bernie supporters either voting third party, voted for Trump or didn't vote. That compares to Clinton supporters in 2008, 30 percent of whom wound up voting for McCain or not voting.
So, defections by supporters of a losing candidate in a contested primary do happen. But there's no reason to panic about how deep the defections will be if Bernie doesn't get the nomination this year. They're not really gonna shoot the dog.

Saturday, February 15, 2020

Hey Liberals, It's Not Just Crazy Trump Voters Who Want Increased Border Security

Polling data consistently show that improved border security is a strong majoritarian position among voters. And that includes Hispanic voters (see below, from Pew data). That is why progressives need to have plans for improved border security--and sound like they mean it. It's a big problem when the Democratic idea that has gotten the most attention this primary season is decriminalizing illegal border crossings. This fits the textbook definition of an unforced error.

Friday, February 14, 2020

No, Radical Policies Won’t Drive Election-Winning Turnout

My new piece in the Washington Post in now available online (it will be in the print edition this weekend). Faithful readers of my FB feed/blog will note some familiar arguments and language. But it's even crisper and clearer this time around!
"No myth is stronger in progressive circles than the magical, wonderworking powers of voter turnout. It’s become a sort of pixie dust that you sprinkle over your strenuously progressive positions to ward off any suggestion that they might turn off voters. That is how Sen. Bernie Sanders (I-Vt.), now the Democratic presidential front-runner, has dealt with criticism that his more unpopular stances — including eliminating private health insurance, decriminalizing the border and covering undocumented immigrants in a government health plan — might cost him the votes he needs to beat President Trump.
Sanders’s explanation of why this is not a problem is simple, and he has repeated it endlessly. When a member of the Los Angeles Times editorial board asked him whether “a candidate as far to the left as you” would “alienate swing voters and moderates and independents,” the senator replied: “The only way that you beat Trump is by having an unprecedented campaign, an unprecedentedly large voter turnout.” Faiz Shakir, Sanders’s campaign manager, adds: “Bernie Sanders has very unique appeal amongst [the younger] generation and can inspire, I think, a bunch of them to vote in percentages that they have never voted before."....
The turnout equation does not necessarily return positive results for a candidate like Sanders. The reverse is more likely. It is truly magical thinking to believe that, in a highly polarized situation, only your side gets to increase turnout. And if the other side turns out in droves, you might not like the results — a warning Democrats would be wise to heed."
Despite what Sanders says, Democrats still have to persuade voters in the middle.