Wednesday, May 29, 2019

Do the Models Say That Trump Will Win?

Well, some models say that but not "the" models. There are many, many election forecasting models so it is not so surprising that you can find a few models that predict a Trump victory. This is particularly the case if you look at so-called fundamentals models that focus on economics and disregard any survey measurements. These are the kinds of models cited by Steven Rattner in his recent Times opinion piece.
The models cited by Rattner say that, if the 2020 election was going to be decided solely on the basis of incumbency and selected aggregate economic measures, Trump would be in very good shape. But I think we kinda knew that already. Put another way, if it weren't for the fact that Trump is so unpopular, he'd be pretty popular.
But wait, aren't these models pretty accurate, despite their seemingly obvious problems?. Nope. Economist Ray Fair's model, for example, is not particularly accurate and has been revised by Fair numerous times to make it come out right (or at least less wrong).
As Nate Silver accurately remarks about these models:
"A lot of smart people don't seem to realize these economic models pretty much suck at predicting elections. They do well when backtested but they're overfit/p-hacked and empirically have done a terrible job of actually predicting elections out of sample.
The ones that use economic data as a prior, and blend it with polling data, do OK and seem to add some value. But the ones based on economy/"fundamentals" alone range from mediocre for the best ones to bordering on junk science for the worst ones."
As an example, Fair's model in 2016 predicted Donald Trump would get 54 percent of the national two-party vote. Whoops.

Tuesday, May 28, 2019

The New Conventional Wisdom on 2018/20 (or At Least It Should Be)

David Jarman at DK Elections has a good write-up of the recently-released Catalist data on the 2018 election and their strategic implications. He doesn't say anything terribly different than what I said about these same data in a recent post, but he does go into a lot more detail than I did. Plus, he has a lot nice simple tables and graphics to get across his points so it's well worth looking at.
Jarman's analysis underscores an important point: the Catalist data are so good and so clear that what they tell us should become the new conventional wisdom about 2018 and the implications for 2020 strategy. Of course, the mobilize-the-base-and-never-talk-to-Trump-voters crowd will still resist. But hopefully these data will convince all but the diehards which way we need to go. As Jarman says:
"[The Catalist release] points out some details that may challenge the conventional wisdom about 2018: Democratic gains in 2018, relative to 2016, were actually the largest in rural areas, not suburban areas, and the bulk of the impact in 2018 came from winning back the votes of 2016 Trump voters, not from the newly activated voters....
This may be the most difficult claim by [Catalist] for people to wrap their heads around....because it contrasts with the way that many Democrats have taken as an article of faith that the way to win going forward is by ramping up young voter and non-white voter turnout..."
But, as Jarman says,the math works out and the data are rock-solid. Time for a new conventional wisdom.
About this website
DAILYKOS.COM
One of the most common intramural arguments that Democrats seem to have is which voters to reach out to in order to try to win close elections: Do you try to turn out “base voters”—usually ...

Friday, May 24, 2019

Surround the Countryside from the Cities!

The Maoist dictum is to surround the cities from the countryside. Democrats need to reverse that dictum and push into the countryside from the cities because, as Will Wilkinson notes in a good piece in the Times, Donald Trump's policies are leaving huge openings for the Democrats in rural areas. This idea is consistent, of course, with the Catalist findings on the 2018 election (reviewed yesterday) that showed Democrats making their largest vote gains in rural areas, essentially due to vote-switching by 2016 Trump voters.
Wilkinson argues:
"President Trump’s feckless trade war is bludgeoning the bottom line of the Republican Party’s reliable rural base. But the party’s disregard for the economic interests of its own constituents goes well beyond barriers to Chinese markets.
Small towns and rural areas, along with some Rust Belt metros, are falling ever further behind booming urban dynamos — leaving many heavily Republican regions in a deepening morass of economic deterioration, joblessness, substance abuse and declining life expectancy. The lower-density places most Republicans call home produce barely half as much wealth as our biggest cities — and it’s showing.
Yet the travails of America’s struggling red regions, and practical ideas about might be done to alleviate them, are barely mentioned in right-leaning policy circles. For example, “The Once and Future Worker,” a widely discussed book by Oren Cass, a former economic policy adviser to Mitt Romney now at the Manhattan Institute, focuses on initiatives to expand employment and wages for American workers but largely neglects the changing geography of economic output and opportunity behind the woes of heartland workers."
He concludes:
"Politically, if Mr. Trump once again chooses divisive culture-war theatrics over an honest attempt to shore up the places that, for now, still prefer Republicans — probably a good bet — Democrats could flip rural House and Senate seats Republicans have long considered “safe.”
This mere possibility could even become likely if only Democratic primary contenders, now seeking favor in rural states, would finally spot the glittering opportunity that Republican misrule has laid at their feet."
Yup.
About this website
NYTIMES.COM
The G.O.P. has left soybean fields littered with $20 bills for enterprising Democratic presidential hopefuls to pick up.

Thursday, May 23, 2019

What Really, Really Happened in the 2018 Election (and Implications for 2020)

I've touted the Catalist estimates that have been released so far as the best estimates we have of voter turnout and preferences by demographic group for the 2018 election and previous elections--and therefore the best tool for helping us understand what really drove the 2018 election results.
So send up the balloons! Yair Ghitza and the good folks at Catalist have just released their final data for 2018, incorporating final precinct-level elections and, critically, individual-level vote history records from voter files around the country. These are some tasty data and as close as we are ever likely to get to a definitive portrait of the 2018 election.
The entire article is up on Medium, with plenty of interesting tables,charts and maps. I urge you to check it out. But here are some items that struck me as particularly important given the post-election strategy debates that have unfolded. .
1. Relative to 2016, the shift toward the Democrats was larger in rural than in suburban areas. This was true among white voters as well.
2. There were big pro-Democratic shifts among both white college (+10) and white noncollege (+7) voters.
3. Turnout was outstanding and the demographic composition of the electorate came remarkably close to that of a Presidential election year. This was due to fewer Presidential dropoff voters and more midterm surge voters.
4. Despite the stellar turnout performance, the overwhelming majority of the Democrats' improved performance came not from less Presidential dropoff and more midterm surge but rather from voters who voted in both elections and switched their votes from Republican in 2016 to Democratic in 2018. When I say "overwhelming" I mean it: Catalist estimates that 89 percent of the Democrats' improved performance came from persuasion--from vote-switchers--not turnout. That's important.
These data imply that 2020 could well be another high turnout election. That should be helpful for the Democrats, who will not and should not stint in their efforts voter mobilization. But the critical role of persuasion will remain, most especially in ensuring that 2018's vote-switchers don't switch back.
About this website
MEDIUM.COM
An Analysis of the Catalist Voter Registration Database

Monday, May 20, 2019

Is Biden Winning Back Obama-Trump Voters?

The short answer to this is "yes" if by that we mean some of these voters are willing to express a preference for Biden over Trump in 2020 trial heats. It's difficult to interpret Biden's significant leads over Trump in states like Michigan, ,Pennsylvania and Wisconsin in any other way.
As Martin Longman notes in a piece on the Political Animal blog, it makes more sense to reason from these polls--which apparently now include polls conducted by the Trump campaign itself--than from articles that quote non-randomly selected working class Trump voters saying how much they still love the President. The latter of course prove nothing other than that such voters exist and the reporter found some.
That said, if by winning we mean in the stronger sense that Obama-Trump-Biden [trial heat] voters are for sure going to vote for Biden over Trump on election day, 2020 if that's the matchup, then of course we can't really say. But it seems promising that at this stage, some of these voters are at least open to going back to the Democrats. As Longman rightly expresses it:
"Ultimately, we cannot know if Biden will be the nominee, nor whether he can win back an appreciable number of Obama/Trump voters, but those aren’t the questions we need to answer right now. First, we need to understand which states are winnable for a Democrat if he or she doesn’t make inroads with white working class voters. Then we need to figure out if there’s an Electoral College path to victory in that scenario. If there is not, or if it looks like a very long shot, finding a challenger for Trump who has “strong support” in these communities will then be vitally important."
About this website
WASHINGTONMONTHLY.COM
Some reports say Biden is beating Trump in the Rust Belt, while others say Trump is holding onto his base.

Thursday, May 16, 2019

Trump Approval in the States and Trump Vulnerability

Nathaniel Rakich has an interesting post up on 538 where he compares Trump's net approval in a state to the partisan lean of a state. He then computes a score showing how much better or worse Trump's net approval is than would be expected from the partisan lean of the state (i.e., if the partisan lean of a state is R+5 you would expect Trump's net approval in the state to also be +5.)
Ignoring Rakich's cute name for the metric (Popularity Above Replacement President or PARP), the interesting thing here is what the metric tells us about swing states and how they may be vulnerable for Trump in 2020.
Unsurprisingly,, the Rustbelt three (Michigan, Pennsylvania, Wisconsin) all look quite accessible for the Democrats. Less so, but still intriguing are Iowa and Ohio. In the southern tier of swingish states, Arizona clearly stands out in Trump vulnerability, followed by Florida and North Carolina and then Georgia and Texas as more of a stretch.
Interesting stuff. It underscores the top three things the Democratic nominee must do in the 2020 general election.
1. Convert Trump disapproval into Democratic votes.
2. Convert Trump disapproval into Democratic votes.
3. Convert Trump disapproval into Democratic votes.
About this website
FIVETHIRTYEIGHT.COM
In recent months, President Trump’s approval and disapproval ratings have been unimpressive, stubbornly stuck around 42 percent and 53 percent, respectively, ac…

Wednesday, May 15, 2019

Biden in Pennsylvania

A new Quinnipiac poll of Pennsylvania voters has Biden looking very, very good, with an impressive 11 lead over Trump in a general election trial heat. On the demographics of his support, I call your attention to not just Biden's huge lead over Trump among white college voters but his comparatively small deficit (-18) among white noncollege voters. Our States of Change data have Clinton losing white noncollege voters in PA in 2016 by 29 points, which accounted for about three quarters of the roughly 6 point swing against her relative to Obama's showing in 2012. If Biden could hold the white noncollege deficit to 18 points in PA in the general election, that would actually be somewhat better than Obama did in 2012 and should be enough to take the state easily.
So a good showing for Biden. He's also the choice of Democratic primary voters by 39-13 over his nearest competitor (Sanders) to be nominee and by 61-6 over his nearest competitor (again Sanders) as the most likely to be able to defeat Trump.

Monday, May 13, 2019

Will the Real Democrats Please Stand Up?

Jonathan Chait attempts to unlock the riddle of why Joe Biden seems to be doing so well, even if he doesn't seem to represent the leftward shift of the Democratic party. Chait points out that the general leftward shift of the party is frequently confused with the viewpoint of its leftmost adherents, which does't really make sense. The whole party can be shifting to the left--I believe that it is--but the median Democrat, even if more left than they once were, can still be pretty different from the aggressively left activists who seem to get most of the attention. Maybe that's why Biden is doing so well, much to the dismay of the Twitterati and the puzzlement of political observers who pay attention to that world. The median Democrat simply wants to beat Trump and a generally left program, even if falls short, of the current AOC-approved laundry list, will be just fine with them.
Of course, that's not what you'll see on Twitter. But, as Chait rightly observes:
"The most important ingredient in the delusion [of a left takeover of the Democratic party] was Twitter. It is hard to exaggerate the degree to which the platform shapes the minds of professional political observers. Part of Twitter’s allure to insiders is that it creates a simulacrum of the real world, complete with candidates, activists, and pundits all responding to events in real time. Because Twitter superficially resembles the outside world’s political debate — it does, after all, contain the full left-to-right spectrum — it is easy to mistake it for the real thing.
But the ersatz polity of Twitter doesn’t represent the real world. Democrats on Twitter skew young and college educated. A study last month found that the Twitter-using portion of the Democratic electorate harbors far more progressive views on everything than the party’s voting base.
One striking example of the disconnect took place earlier this year in Virginia. An old medical-school yearbook showed Ralph Northam, the state’s Democratic governor, in a picture featuring a blackface costume and Ku Klux Klan robe and hood. If you followed the debate on Twitter, as nearly all political reporters did, Northam’s resignation was simply a given. The debate turned to when he would step down, who would replace him, and what other prominent people would have career-ending blackface yearbook photographs.
Virginians, however, were split in ways the political elite would never have guessed. Whites and Republicans favored his resignation, while African-American voters believed, by a 20-point margin, that Northam should not resign."
So don't believe what you see or hear on Twitter. Biden doesn't and that appears to be serving him very well. Other candidates should take note.
About this website
NYMAG.COM
His apparent resurrection from relic to front-runner has illustrated a chasm between perception and reality about the party’s leftward shift.

Sunday, May 12, 2019

How Vulnerable Is Biden on the Green New Deal?

Biden is already taking a lot of incoming from his Democratic rivals and GND partisans on the inadequacy of his climate plan, despite few details being currently available. What little we know, as summarized in a Reuters report:
"Democratic presidential hopeful Joe Biden is crafting a climate change policy he hopes will appeal to both environmentalists and the blue-collar voters who elected Donald Trump...carving out a middle ground approach...
The backbone of the policy will likely include the United States re-joining the Paris Climate Agreement and preserving U.S. regulations on emissions and vehicle fuel efficiency that Trump has sought to undo...[T]he policy could also be supportive of nuclear energy and fossil fuel options like natural gas and carbon capture technology, which limit emissions from coal plants and other industrial facilities."
Despite all the heavy breathing from his opponents and miscellaneous advocates, this doesn't sound so terrible to me as an opening bid. Of course, we'll see what the details of the plan are but I am not too worried that Biden is not unabashedly endorsing a huge Green New Deal of the AOC-Markey variety.
In fact, it's probably a good idea not to do that. Polling indicates that voters are very supportive of clean energy and are open to some kind of Green New Deal, but the details of such a plan are clearly going to matter a lot to its saleability in a general election campaign, not to mention actual feasibility as policy. So it's fine the Biden is apparently leaving himself a lot of room to maneuver in this area.
In the end, I suspect that will not only be good for Biden and Democratic prospects in 2020, but also for actual progress on a program that can combat climate change, even if winds up under a different name. As the spectacularly effective leader Deng Xiaoping put it, "Who cares if a cat is black or white, so long as it catches mice?"
About this website
NYTIMES.COM
A report that Mr. Biden is seeking a “middle ground” on climate change has sparked attacks from progressive Democratic candidates.

Friday, May 10, 2019

The Demographics of Biden's Support

This is a very nice table released by ALG Research showing Biden's support by Democratic subgroups. His support is strongest among older voters, noncollege grads, moderate/conservative Democrats and, interestingly, voters of color and women. The only group where he doesn't have a lead is very liberal Democrats--maybe a fifth of Democratic voters.
And of course he's just far ahead of the other Democratic candidates period. That includes in the three early voting states of Iowa, New Hampshire and South Carolina..He also leads Trump nationally by 7 points, by 13 points in Michigan, and is the only Democrat who leads Trump in Arizona or Nevada.
The data say Biden looks pretty strong. But he can he bring the party together in an effective way to defeat Trump and move the Democratic agenda forward? As Paul Starr puts it: "The Democratic Party today is both a coalition and a movement, and the presidential nominee has to lead the movement while holding the coalition together." Can Joe do it? Starr has some ideas about how he can in his latest American Prospect column. I recommend it.

Thursday, May 9, 2019

How Far Left Is Too Far Left?

Public opinion and recent electoral results suggest American voters are ready for significant change in a leftward direction. Consequently, the Democratic nominee in 2020 is likely to offer a strongly progressive program to the electorate.
Common sense suggests, however, that there are limits to how far left this program should go. Programs that lack public support should be avoided and programs preferred that generate strong support among not just among the base Democratic constituencies like blacks and Latinos but also among white college graduates and ideally have significant purchase among white noncollege voters as well.
The latest Quinnipiac poll provides some insight along these lines. The Democratic policy ideas tested in the poll included several that did not garner majority public support, including making all public colleges free to attend (45 percent), a marginal tax rate of 70 percent on income over $10 million (36 percent) and allowing current prisoners to vote (31 percent).
But there were a couple of ideas that got strong support. One was forgiving up to $50 thousand in student debt for households making under $250 thousand a year. Overall 57 percent support included very high support from blacks and Latinos but also solid support from college and noncollege whites.
Support was even stronger for an annual wealth tax of 2 percent on those with over $50 million. Three-fifths of voters supported this with blacks, Latinos and college whites all strongly in favor and even noncollege whites favoring the proposal by 15 points.
Also worth mentioning here is the latest Kaiser poll which showed Medicare for All getting only 37 percent support if it eliminated private insurance, but Medicare for all who want it, while retaining the private insurance option, drawing lopsided support from not only blacks and Latinos but college and noncollege whites as well.
So there is left and then there is too far left. Democrats who want the best chance of beating Trump and the most robust possible political coalition would be wise to choose the former not the latter.
POLL.QU.EDU
Biden Surging Among Democrats In Presidential Race, Quinnipiac University National Poll Finds; U.S. Voters Support Wealth Tax, Oppose Free College