There's no counting the number of times the media have asked: Will this be the thing that drives Donald Trump's supporters away from him? Is this finally the time?

During the campaign, it was incessant: the times he'd insult his opponents, the Access Hollywood tapes, the latest foray into "political incorrectness." Once he was elected, the focus shifted to his appointments and proposals: choosing Goldman Sachs executives for his senior team and cabinet, bombing a Syrian airfield, proposing a budget that cuts social program spending.

Whether the president's decision is good or bad, the action justifiable or appalling, the question is always, always posed: Won't this drive away his voters?

Coming from the polling world, I've long since stopped expecting that any one thing is "the thing" that will dramatically shift Trump's numbers downward. And with exciting new research out on just who comprises Trump's coalition, I believe that part of the resilience of Trump's numbers comes in large part because not all Trump voters are alienated by or attracted to the same things about the president.

For all the constant churn of the news cycle and the endless stream of outrage, backlash, and bluster, Trump's job approval numbers have stayed surprisingly stable since his inauguration. From around 44 percent during his "honeymoon" period, Trump's approval numbers are still around 40 percent. For comparison, Barack Obama's approval rating started in the mid-60s, and by late June of his first year, it was approaching the mid-50s. (Still better numbers overall than those enjoyed by President Trump, to be sure.)

Part of this is because of the calcification of views of Trump among Democrats: Trump hasn't lost ground with the Left because he never had ground with them to start with. Obama's numbers fell by a slightly larger amount over his first few months because he enjoyed much more support right at the start from Republicans, support that eroded quickly.

But I suspect part of this has to do with the ideological diversity and varied interests of Trump's coalition. A new study out from Democracy Fund's Voter Study Group – a research coalition I was honored to advise – covers a great deal of ground explaining and understanding the who, what, why and how of the 2016 election.

In the study Cato Institute scholar Emily Ekins lays out typology of the Trump voter, finding five clusters of voters in his coalition. In a sense, most of these clusters are represented by at least a handful of Trump White House senior advisers. The diverging views of these advisers look like a source of conflict, but they also make the Trump administration look like Trump's electorate.

"Staunch conservatives" are the largest cluster, making up 31 percent of Trump's voter base. These are traditional Republican types, favoring conservative social and economic policy. Reince Priebus, the career party operative, fits this mold.

The second-largest group is the "free marketeers," at 25 percent. These are your Wall Street types, those with fairly moderate social views and who tend to be pro-free trade and pro-immigration, embodied in the White House by either Ivanka Trump or Gary Cohn, the Goldman Sachs alumnus.

Behind them come the "American preservationists," who are 20 percent of Trump's coalition. Staunch conservatives and free marketeers are fairly typical Republicans, while the American preservationists are far less reliably a part of a GOP coalition. These folks are in many ways economically progressive, with what Ekins calls "a nativist and ethnocultural conception of American identity." Think of them as the Steve Bannon wing of the coalition.

In roughly equal number are simply the "anti-elite," 19 percent of the coalition, holding more moderate immigration and racial views than the American preservationists, but with no less anger at institutions and a sense that Washington isn't working or looking out for people like them. An additional 5 percent, "the disengaged," feel detached generally and do not have many strong preferences.

When Trump issued the order to bomb an airfield in Syria that had been the launching point for chemical weapons attacks, some wondered: Would this alienate "the Bannon wing"? Trump's decision to pull out of the Paris Accords: Was it a blow to "the Ivanka wing"?

And yet after each of these things, Trump's numbers don't move much. Bombing Syria may have gotten Trump backlash from the alt-right, but he continues to fight angrily on Twitter for his travel ban. Pulling out of Paris Accords may have been dismaying to some of the more moderate elements of Trump's coalition, but hey, there's still tax reform to be had.

The things that frustrate one piece of Trump's coalition often endear him to or embolden another wing of the coalition. If he hovers around 40 percent over the months, it's not that he's holding his whole base steady — it's because he's losing a few supporters with each move but gaining a few more. There's little he could do to please his whole base but also little he could do to anger his whole base.

Of course, he needs to deliver. All the angry tweets in the world won't hold the coalition together if the various factions feel the president hasn't scored any wins for them as time goes on. But for the moment, next time you see the latest outrageous thing the president has tweeted and wonder "Will this be the thing that drives his voters away?" remember that "his voters" are not all the same.

Kristen Soltis Anderson is a columnist for The Washington Examiner and author of "The Selfie Vote."