Campaign 2020 shapes up as a referendum on the incumbent. Yet if recent data is any guide, Americans have made up their minds about Trump, and nothing between now and Election Day - not actual events, and certainly not the spin his political opponents put on those events - is likely to change that.
Much has happened between April 1, 2017, and April 1, 2019. A partial list: Trump's firing of FBI Director James Comey, a violent white-supremacist march in Charlottesville, devastating hurricanes in Puerto Rico and Florida, huge tax cuts, the Brett Kavanaugh Supreme Court confirmation hearings, family separation on the U.S.-Mexico border, a 35-day government shutdown and an embarrassing Trump-Putin news conference in Helsinki.
After that tornado of events passed, 42.3% of the public approved of Trump's job performance, in an average of polls reported by FiveThirtyEight, the political website; 52.8% disapproved. Two years previously, on April 1, 2017, the figures were 41.5% approve, 52.5% disapprove.
To repeat: 24 months of tumult produced essentially no change in Trump's standing with the public. Strictly speaking, it improved. His net disapproval rating had shrunk slightly, from 11 points to 10.5. (Another leading "poll of polls," produced by RealClearPolitics, confirms this trend.)
Looking back, there have only been two periods of significant malleability in public views of Trump. The first big change was a 20+ percentage-point surge in Trump's personal favorability rating (measured by RealClearPolitics) between the time he started his run in mid-2015 and the end of 2016. That reflected a pro-Trump shift among previously skeptical Republicans.
The second came in the first three months of his presidency, when the public went from being evenly divided on his job performance to disapproving of it by double digits. The last remaining Democrats and independents willing to give Trump a chance apparently recoiled from his belligerent tweets and from such policies as his initial attempts to limit Muslim arrivals in the United States.
Other than that, Trump's approval rating has remained mostly in the low-to-mid 40s, while his disapproval number has hovered in the low-to-mid 50s. According to FiveThirtyEight, Trump's numbers have been much less volatile over the first 800 days of his presidency than those of any president since Harry S. Truman.
Only Barack Obama comes close to Trump in this respect, which is no accident. Like Obama, who could count on steady support from a core constituency - black people - Trump enjoys near-unconditional backing from white evangelicals. In both cases, the effect was to help build a floor below which the approval rating cannot fall.
The United States is now thoroughly sorted into two camps: a large minority of people who are pretty much pro-Trump, and a majority that are pretty much anti-. Obviously a major downturn in the economy or the outbreak of a war could change all that, but neither appears likely today. A fair assumption is that Trump's share of the popular vote in 2016, 46.1%, represents his likely total in 2020, give or take a percentage point.
If that assumption holds, and Democrats manage to distribute their share of the vote more widely than they did in 2016, Trump would be unlikely to win the Electoral College again.
The 2020 race looms as a base-mobilization contest, in which victory will likely belong to the tribe - er, party - that motivates its people to turn out. That is to say, it will be the same kind of election that 2016 was, only this time the Democrats will be fully aware of that going in.
But the bad news for Democrats is that, beyond a certain point, Trump is impervious to negative campaigning. As the past two years show, there is only so much Democrats can do to push his support levels down, or to push their candidate's up, for that matter.
The key variables appear to be who the Democrats choose as their nominee, and how they choose him or her. They can select someone so far to the left that frightened moderates stay home, vote for a third-party candidate or even pick Trump as the devil they know.
Or the Democrats can attack one another so furiously in the crowded primary contest that they fatally divide and demoralize the anti-Trump majority.
Trump's capacity for beating the Democrats is limited, but the same cannot be said of the Democrats' capacity for beating themselves.
Sign up for the daily JWR update. It's free. Just click here.