If you decrease your expectations for the polls you will be much better off — but that doesn't mean they are useless, it means most people have been underrating potential error and misunderstanding exactly how (in)accurate they can be
For example, we assigned a 5% chance to Biden's actual vote margin in Wisconsin. That was probably too low (and we have written about why.) But the media reaction to that miss seems closer to what you'd want to see with a 1-in-100 or 1-in-1,000 miss. It's all miscalibrated.
The cool thing about blogging is that, over time, you get enough content to just footnote all your tweets
Anti-majoritarianism has gained a firm foothold on the American right over the last few years. Polls offer one partial remedy to our historically-biased electoral institutions, and the conservative outlets that are increasingly hostile to the general will. gelliottmorris.substack.com/p/seeing-the-p…
I understand this is a tough argument to make — the deck is stacked against the majority in ways that simply sharing (good!) polling data isn’t going to fix. But my hope is with the right volume of coverage and a good attitude in Washington, we can come to see polls differently.
I have been blogging a lot recently about how we see and use the polls in non-electoral contexts, and if you’re into polls I’d really like if you put your name on my email list so I can tell you when my book on the same subject comes out next year:
Not sure people have really noticed this: the share of Americans favoring a legal path to citizenship for undocumented immigrants has just hit a 4-year low civiqs.com/results/immigr…
The only legitimate outcome in a democratic republic is for the majority to get what it wants the vast majority of the time. US institutions violate that principal. The minority’s only defense for this is “well that’s good for me,” which is a bad defense!
The big reason why Americans have to obsess over protections for the political minority is because our institutions are so insulated against political accountability. If we had parliamentary democracy, or more proportional legislative chambers, the <50% would be at much less risk
(This is partly because the government is more responsive, but also because you don’t get factions that are committed to violating fundamental liberal rights to hold onto power)
If Trump's gains among Hispanics, esp conservatives, between 2016 and 2020 came in large part from an increasing focus on income growth/other economic gains, we could view that in part as an incumbency bonus — part of which might flow back to Biden in 2022/2024.
Biden will continue to face challenges of broader ideological sorting among conservative non-whites, but Dem gains among moderate and liberal whites — who make up a much larger share of the electorate — is an underrates story for 2020, both for POTUS and Dems taking the Senate
The shift among Latinos is a popular topic among media elites in part because it seems unfathomable to them that non-whites (or anyone really) would vote for Trump, but this stems from a broader misunderstanding of voter psychology and misses the relative importance of the change
Ackshually, the only “good” “compromise” is to retro-retrocede Alexandria and Arlington back into DC then grant residents statehood, that way Rs have a higher chance of winning senate seats in VA but DC residents get what they want