Is it time for Democrats to panic? Well, isn't it always?

I kid. But if you're a keen observer of the news, you'll notice that the story about the upcoming midterms has been shifting from "Blue wave coming!" to "Hang on, maybe not!" As The Washington Post noted on Monday, "After months of confidence that public discontent with President Trump would lift Democrats back to power in Congress, some party leaders are fretting that their advantages in this year's midterms are eroding amid a shifting political landscape."

President Trump regularly cites the widely-mocked Rasmussen poll to show how high his approval ratings are. "The president is strong in these states" in the Midwest, says Speaker of the House Paul Ryan. "He's an asset." Analyst Charlie Cook sees "glimmers of hope for the GOP." You've probably heard that Trump's approval has risen, and the generic ballot test has tightened.

No one knows for sure what will happen, of course. But there's a powerful impulse at work here: the desire for a new story.

That doesn't necessarily mean anyone is insincere in what they're observing or how they're analyzing it. But there's a reason the word "news" comes from "new." All of us, particularly those of us in the news media, are always on the lookout for what isn't like it used to be. Write (or read) one story enough times — like the story that Democrats are headed for a big victory in the fall — and you start itching for something new, and focusing on the indicators that suggest that things have actually changed.

So let's try to put things in perspective. Has President Trump's approval risen lately? The answer is yes — but not by much. If you look at FiveThirtyEight's aggregation of polls, you'll see that a month ago, Trump was at 40 percent, and now he has rocketed all the way up to ... 42 percent. In fact, his approval has been remarkably steady over his entire time in office, seldom shifting more than a few points from 40 percent in one direction or the other.

That may be surprising, given the regular stream of lunacy emanating from the White House. But it shows the durability of partisan attachment: Even a presidency as alternately catastrophic and comical as this one won't drive away more than a handful of voters from the president's party.

At the same time, however, Trump is indeed different from presidents before him. Ordinarily, a president overseeing an economy in which unemployment had fallen below 4 percent would be basking in the affection of most Americans. Yet Trump's approval has barely budged from where it has always been, even as the recovery that began under Barack Obama has deepened.

Republicans are hoping that Trump is so different that he can become the exception to what for them is a very frightening rule: When the president's approval is below 50 percent, his party doesn't just lose seats in the midterms, it loses big.

In 2014, Barack Obama was at 42 percent approval in Gallup's polling just before the midterms, exactly where Trump is now, and his party lost only 13 seats in the House, much less than the 23 seats Democrats need to take control. But that was in large part a product of how many seats they lost four years before, since there just weren't many remotely competitive seats controlled by Democrats left over. In 2010, Obama was at 45 percent approval — and Democrats lost a net of 63 seats.

Other elections show a similar pattern. In 2006 George W. Bush was at 38 percent, and Republicans lost 30 seats. In 1994, Bill Clinton was at 46 percent, and Democrats lost 53 seats. While those suggest a wide range of outcomes — and many other factors that play into the results — there's an apparently immutable law underneath. Unless something extraordinary happens that upends the entire political world and makes the president unusually popular, in his first midterm the president's opponents are going to be highly motivated, and his party is going to lose big.

It's always possible that something extraordinary could indeed happen between now and November. But it would have to be of a monumental scale. The only recent examples are the Sept. 11 attacks, which boosted Bush's popularity for 2002, and impeachment, which ironically boosted Clinton's popularity for 1998. It's awfully hard to imagine what would suddenly improve Trump's standing so much that the voters who currently can't stand him would suddenly decide he's doing a good job and needs a Republican Congress to help him out.

Of course, there's plenty of time for Democrats to screw things up, which they've shown a talent for in the past. But notwithstanding a jump here or there in one poll or another or a race here or there where a questionable candidate wins the primary, you still wouldn't want to be a Republican running for office this year. And that wave is probably still on it way.