One of the downsides of serving President Trump is that, occasionally, you’ve got to toe the line. The president tweets a conspiracy theory, and you’re put in the position of somehow squaring it with reality.

On Sunday, FEMA Administrator William “Brock” Long starred in his episode of this long-running serial. Appearing on the Sunday shows, he was asked to explain Trump’s untrue tweets suggesting the Puerto Rican government’s official death toll after Hurricane Maria — 2,975 — was wrong and “was done by the Democrats to make me look as bad as possible.”

Here’s a recap of what Trump has said on the subject:

Long argued the death toll might, in fact, have overshot reality and cast the response in an unfair light. At one point, he even suggested Trump was unfairly blamed for deaths resulting from spousal abuse after the hurricane. But he was unable to account for Trump’s allegation that Democrats somehow cooked the books.

Below are Long’s exchanges with “Meet the Press” host Chuck Todd and “Face the Nation” host John Dickerson, with annotations in yellow.

‘Meet the Press’

TODD: The president has been disputing the death toll, multiple tweets. He said, “Over many months, it went to 64 people, then like magic, 3,000 people killed.” Believe it or not, Mr. Long, the White House put out a five-page backgrounder citing other death toll numbers that were less than the one that the governor of Puerto Rico has accepted. 2,975 people died as a result of Hurricane Maria or impacts from Hurricane Maria, according to the Puerto Rican government. Does FEMA accept that number?

LONG: So, the numbers are all over the place. FEMA doesn’t count deaths. And like, if you take what’s going on with Florence: The deaths that are verified by the local county coroners are the ones that we take. Now, what we do offer are funeral benefits after a disaster for those that are eligible. And so, those are some of the numbers that you can put forward that can be cross-referenced with any of the numbers that are out there. But here’s the thing, you know, these guys are so dedicated. They work around the clock. One death is a death too many. You know, and we —

TODD: But then why does it matter —

LONG: — and we push forward.

TODD: Why is the White House so concerned about the difference between 3,000 deaths and, say, another report that might have had it at 1,800 deaths? I mean, you’ve said yourself, it doesn’t matter, but the White House believes it matters. Why?

LONG: Well, I’ll tell you this, you know, one thing about President Trump is, is that he is probably the one president that has had more support for what goes on back here. And I think he’s defensive because he knows how hard these guys behind me work day in and day out for a very complex situation. And it’s frustrating. Those studies, the Harvard study was done differently than the George Washington study, or this study or that study, and the numbers are all over the place. And where the 65 —

TODD: Were any of them —

LONG: Hold on. Hold on.

TODD: Well, is it fair — I mean, he said Democrats did it to make him look bad. Do you believe any of these studies were done to make the president look bad?

LONG: Well, I mean there’s, I don’t think the studies, I don’t know why the studies were done. I mean, I think what we’re trying to do, in my opinion, what we’ve got to do is figure out why people die, from direct deaths, which is the wind, the water and the waves, you know, buildings collapsing, which is probably where the 65 number came from. And then there’s indirect deaths. So, the George Washington study looked at what happened six months after the fact. And you know, what happens is — and even in this event, you might see more deaths indirectly occur as time goes on because people have heart attacks due to stress. They fall off their house trying to fix their roof. They die in car crashes because they, they went through an intersection where the stoplights weren’t working. You know the other thing that goes on, there’s all kinds of studies on this that we take a look at. Spousal abuse goes through the roof. You can’t blame spousal abuse, you know, after a disaster on anybody. And the president’s very passionate about the work we’ve done. He’s been incredibly supportive of me and this staff. He’s actually — I bet he’s probably the only president that’s held two Cabinet-level meetings, brought his entire cabinet to this agency to show support. They come through this agency every day. And he’s, he’s, he’s very supportive, which is exactly what FEMA needs. There’s just too much blame going around, and we need to be focused, Chuck, on what is Puerto Rico going to look like tomorrow.

‘Face the Nation’

DICKERSON: The interruption of medical care that you were just talking about was responsible, I believe, for 47 percent of the fatalities in Katrina and is a big part that number that's been disputed this week about Puerto Rico, the 3,000 number.

So, the president said that 3,000 number didn’t exist, that they didn’t die. So how is it true that you’re preparing for interruption of medical care in Florence, but the president says people who died as result of interruption in medical care in Puerto Rico are not worth counting?

LONG: Well, you know, look, these studies are all over the place.

The Harvard study was done differently, studied a different period of time vs. the George Washington study. There's a big discrepancy, whether it's direct deaths or indirect deaths.

If you look at the root cause of any problem is, one, around here, one death — these guys know one death is a death too many. And we work every day to make sure that we try to prevent that.

But if you want to get into Puerto Rico from the standpoint of what needs to happen next, you have got to fix aging infrastructure that wasn’t ready to support the commonwealth before the storm hit, and when they were blown out, and the infrastructure is blown out, it exponentially causes problems on the back end.

DICKERSON: But the reason it's so important, obviously, if you figure out how people died last time, you can keep it from happening again.

You say the numbers are all over the place. But the numbers are more than zero, which is what the president said. He said the deaths didn't happen.

I guess my question is this. The G.W. report, as you mentioned, again, the bulk was from interruption of medical care, which you're trying to take care of in Florence. They interviewed people from FEMA to come up with that number. So who is right, the president, who says those deaths didn't happen, or the FEMA officials who helped G.W. put together that report?

LONG: Yes, I don’t know who they interviewed within my agency. They may have looked at funeral benefits to help calculate whatever number.