Ron Brownstein does a nice job laying out the forces and counterforces that will determine the outcome of the 2020 election. They are:
"The three biggest challenges looming in 2020 for Trump, many analysts agree, are:
* The recoil from his definition of the Republican Party in white-collar suburbs, including many that previously leaned toward the GOP.
* A feedback loop in which his efforts to mobilize turnout among his core supporters are producing an offsetting turnout surge among key Democratic groups, particularly African Americans.
* An unremittingly confrontational personal style that appears to be alienating a broad swath of female voters, including some of the non-college white women who helped drive his 2016 victory. That behavior was exemplified by Trump's tweet last week attacking former US Ambassador to the Ukraine Marie Yovanovitch in bitterly personal terms.
Trump's principal political assets on the other side of the ledger are his success at consolidating and energizing the Republican base and deepening the GOP's dominance among white voters who live outside of major population centers, identify as evangelical Christians or lack college degrees, especially the men in each of those groups."
If I had to pick a demographic that I think will determine the 2020 result in the last instance, I would be tempted to pick white noncollege women. If his evident softness among this group translates into a lack of vote support next November, I think it'll be very hard for him to win.
"In Wisconsin polling by the Marquette University law school, Trump's approval rating among non-college white women averages just 42% through his presidency; the latest Muhlenberg College survey in Pennsylvania found that he led Democratic Joe Biden among them by just 5 percentage points (after beating Hillary Clinton by 20 points with them there in 2016, according to the exit polls). Recent state surveys by the Kaiser Family Foundation and Cook Political Report likewise put Trump's approval among non-college white women at just 42% in Michigan, 43% in Wisconsin and 46% in Pennsylvania. Nationally, an average of the weekly polls conducted since July by the Nationscape project, launched by the Democracy Fund and UCLA political scientists, found that Trump's approval among non-college white women who are not evangelical Christians -- who account for most non-college white women in the Rust Belt -- stood at just 41%."
But it's still way early. Keep your eye on the trends mentioned by Brownstein but remember: it's not just the trends; it's how they net out. That's the big and, at this point, unanswerable question.