FiveThirtyEight

A widely held tenet of the current conventional wisdom is that while President Trump might not be popular overall, he has a high floor on his support. Trump’s sizable and enthusiastic base — perhaps 35 to 40 percent of the country — won’t abandon him any time soon, the theory goes, and they don’t necessarily care about some of the controversies that the “mainstream media” treats as game-changing developments.

It’s an entirely reasonable theory. We live in a highly partisan epoch, and voters are usually loyal to politicians from their party. Trump endured a lot of turbulence in the general election but stuck it out to win the Electoral College. The media doesn’t always guess right about which stories will resonate with voters.

But the theory isn’t supported by the evidence. To the contrary, Trump’s base seems to be eroding. There’s been a considerable decline in the number of Americans who strongly approve of Trump, from a peak of around 30 percent in February to just 21 or 22 percent of the electorate now. (The decline in Trump’s strong approval ratings is larger than the overall decline in his approval ratings, in fact.) Far from having unconditional love from his base, Trump has already lost almost a third of his strong support. And voters who strongly disapprove of Trump outnumber those who strongly approve of him by about a 2-to-1 ratio, which could presage an “enthusiasm gap” that works against Trump at the midterms. The data suggests, in particular, that the GOP’s initial attempt (and failure) in March to pass its unpopular health care bill may have cost Trump with his core supporters.

These estimates come from the collection of polls we use for FiveThirtyEight’s approval ratings tracker. Many approval-rating polls give respondents four options: strongly approve, somewhat approve, somewhat disapprove and strongly disapprove. Ordinarily, we only estimate Trump’s overall approval and disapproval. But we went back and collected this more detailed data for all polls for which it was available, and then we reran our approval ratings program to output numbers for all four approval categories instead of the usual two. Here are Trump’s strongly approve and somewhat approve ratings from shortly after the start of his term through this Tuesday:

After a slight uptick in the first two to three weeks of his term, Trump’s strong approval ratings have headed downward. But it hasn’t been a steady decline. Instead, they fell considerably from about 29 percent on March 6 — when Republicans introduced their health care bill — to around 24 percent on April 1, shortly after the GOP pulled the bill from the House floor. They then remained stable for much of April, before beginning to fall again this month after the reintroduction (and House passage) of the health care bill and after Trump fired FBI director James Comey on May 9. As of Tuesday, just 21.4 percent of Americans strongly approved of Trump’s performance.

By comparison, 45 percent of Americans strongly approved of President Obama’s performance as of April 2009, although Obama’s strong approval numbers would fall considerably over the course of his term — to the mid-to-high 20s by the midterms and to the high teens by 2014.

The share of Americans who somewhat approve of Trump’s performance has actually increased slightly, however, from about 16 percent in early February to 17.9 percent as of Tuesday. In part, this probably reflects voters who once strongly approved of Trump and who have now downgraded him to the somewhat approve category. (Trump’s strongly approve and somewhat approve numbers have been inversely correlated so far, meaning that as one has risen, the other has tended to fall.) A potential problem for Trump is that in the event of continued White House turmoil, the next step for these somewhat approve voters would be to move toward disapproval of the president.

The number of Americans who strongly disapprove of Trump has sharply risen since early in his term, meanwhile, from the mid-30s in early February to 44.1 percent as of Tuesday. In most surveys, Trump’s strongly disapprove rating exceeds his overall approval rating, in fact.

The bulk of the increase in Trump’s strong disapproval ratings came early in his term, over the course of late January and early February. It’s possible that this was partly a reaction to Trump’s initial travel ban on immigrants from seven predominantly Muslim countries, which was the biggest news of Trump’s first few weeks in office. But presidential disapproval often rises in the first month or so of a president’s tenure as voters who initially give a new president the benefit of the doubt find things to dislike in his performance.

Meanwhile, the share of Americans who somewhat disapprove of Trump has been small and fairly steady throughout his term, usually averaging around 10 or 11 percent. It was 11.6 percent as of Tuesday.

During last year’s presidential primaries, Trump received about 14 million votes out of a total of 62 million cast between the two parties, which works out to 23 percent of the total. So perhaps it’s not a coincidence that 20 to 25 percent of the country still strongly supports Trump; they were with him from the start.

But 20 to 25 percent isn’t all that large a base — obviously not enough to win general elections on its own. Instead, Trump won the White House because most Republicans who initially supported another GOP candidate in the primary wound up backing him in the November election. Trump has always had his share of reluctant supporters, and their ranks have been growing as the number of strong supporters has decreased. If those reluctant Trump supporters shift to being reluctant opponents instead, he’ll be in a lot of trouble, with consequences ranging from a midterm wave against Republicans to an increased likelihood of impeachment.

So while there’s risk to Democrats in underestimating Trump’s resiliency, there’s an equal or perhaps greater risk to Republicans in thinking Trump’s immune from political gravity.

If you look beneath the surface of Trump’s approval ratings, you find not hidden strength but greater weakness than the topline numbers imply.

Harry Enten and Dhrumil Mehta contributed to this article.