The 2018 midterm elections are a mere 557 days away, and Democrats are feeling quite confident. There are many reasons to think that we could be headed for a sweep to rival what they accomplished in 2006 when they took back both houses of Congress.

But at least one Democratic pollster is warning his fellow Dems that they could actually be heading for terrible disappointment, and he raises some intriguing questions.

We’ll get to those in a moment, but first we should acknowledge all the reasons Democrats have to feel good about 2018. The first is the widely accepted historical pattern that the opposition party does well in midterm elections, as voters express their displeasure at the president by turning out to vote. In addition, unless things are going unusually well or some historical anomaly intervenes — the way 9/11 affected the 2002 election or Bill Clinton’s impeachment shaped 1998 — voters in the middle often blame the president for whatever they’re dissatisfied with, increasing the ruling party’s headwinds.

Today we learned that gross domestic product growth in the first quarter of this year was a mere 0.7 percent, and though there’s no way to know what it will be a year and a half from now, chances are strong that President Trump won’t bring us so much winning we’ll be tired of winning. His elaborate promises will be all but impossible to fulfill, especially given the fact that by most measures the economy was doing quite well when he took office, with incomes rising and unemployment below 5 percent. He might be fortunate enough to see things stay as they are, but even that could wind up looking like failure.

Meanwhile, all of his weakness are on full display, and his approval rating sits at around 40 percent, lower than any of his predecessors in the history of polling at this stage of their presidencies. Perhaps most important, his election has led to an absolute eruption of grass-roots energy and activism on the left, with local Democratic Party organizations being overrun with new volunteers and groups such as Indivisible sprouting up chapters at a fierce clip. While many potentially strong Republican candidates are deciding to sit next year’s elections out to wait for a more welcoming landscape, Democrats are lining up to run: Emily’s List says that while in 2016 it was contacted by 900 women looking for help and advice in running for office, this year 11,000 women have reached out to them already.

So that’s the case for Democrats to be optimistic. What’s the other side of the argument? Democratic pollsters Allan and Sheri Rivlin have been going around Washington showing Democrats a PowerPoint presentation to try to convince them that they need to revamp their economic message, which few would take issue with. But more controversially, they assert that the assumption that Democrats will do well because of the “midterm curse” — the notion that the president’s party will inevitably lose seats in the midterm — is dangerously naive.

“The midterm curse is this year’s version of the ‘blue wall,'” Allan Rivlin told me, referring to the Clinton campaign’s belief that their hold on Pennsylvania, Michigan and Wisconsin would guarantee their victory in 2016.

The beginning of the Rivlins’ case for caution is their observation that in recent history Democrats have never won 20 or more seats to take back the House in the first midterm of a new Republican president, and this year they need a net gain of 24 seats to do so. Most of the truly enormous midterm waves have favored Republicans, especially those of 1994 (where they gained 54 seats) and 2010 (63 seats), both at a Democratic president’s first midterm.

“Democrats do poorly in midterms,” Rivlin argues. “Republicans are rarely on the losing side of this.” It’s partly because Republican voters — older, whiter, more affluent — are more likely to turn out in any election while many Democrats don’t bother showing up in midterms. But Rivlin is especially concerned with Democrats’ lack of a core economic message, since the economy is usually voters’ most important issue. “We think we have an economic message,” he says, “but we don’t.”

And this has been a problem for Democrats for some time now. “The voters have been expressing more trust in Republicans to get the economy going and create jobs than Democrats,” Rivlin notes. “This was true in 2010, 2014 and 2016, which is why in all three of those elections I was the biggest pessimist I knew.”

What Democrats lack is a message on economics that can pass what he calls “the Listerine test.” Listerine had what Rivlin describes as a nearly perfect message: “Listerine kills the germs that cause bad breath.” Eight words that describe the problem, the solution and how it works.

The Republican message on the economy passes this test. It’s simple, easy to understand, and explains both every economic problem you could think of and what their solution is: Government is the problem, so if we cut taxes and cut regulations, the economy will blossom.

Democrats look at that message and cry, “But it doesn’t work!” We’ve tried it before, and the Trump administration is about to try it again, but we already know how it’s going to turn out. Prosperity for all is not on its way once we cut taxes and allow more pollution. But here’s the problem, according to Rivlin: Democrats have convinced themselves that the Republicans’ message is bogus, but average voters still respond quite positively to it, no matter how many times the policies fail in the real world.

And what do Democrats say in response? “We say we care about jobs and the middle class, and then we list three policies.” Rivlin notes that the specific policies Democrats advocate, such as a higher minimum wage, equal pay for women, or infrastructure investments, test very well in polls. But they don’t tell a broad, encompassing story about the entire economy. “We don’t have this overarching answer to the question: What’s your diagnosis of the problem, and what’s your solution? We don’t share that with an economy of words, and Republicans do.”

I’d contend that this is a problem Democrats face more in 2020 than in 2018. What’s more important than anything else in midterm elections is which side is riled up and angry, and there’s no question where the anger is right now. Yes, Democrats face a natural disadvantage because their base voters don’t turn out at the same rates the Republican base does, but their highest priority for 2018 should be about keeping that base motivated and engaged.

That doesn’t mean you can’t do both things at the same time, however. “A hundred percent of our attention is in pointing out Trump’s weaknesses,” Rivlin says. “We need to take some of that attention to addressing our own weakness.”

Unfortunately, he doesn’t yet have the perfect answer to the what the Democrats’ Listerine test should be. He does have a collection of ideas he says Democrats need to explore, centered around things such as innovation, strengthening the middle class and reducing inequality. But his argument is that Democrats need a focused project to work through these ideas and determine which one best describes the large number of policies they’re already advocating and will be most persuasive to the public.

In recent months, Rivlin has given this presentation to dozens of Democrats on Capitol Hill, in think tanks and among party professionals. When I asked him what the response has been, he said that at first everyone enthusiastically agrees. But that doesn’t last. “Then I lose contact with them, because Trump does something that requires an all-hands-on-deck response,” Rivlin says. “Everybody gets snapped back into oppose-Donald-Trump mode.”

That’s understandable. But eventually — if not in the next year then certainly before they mount another presidential campaign — it’s a problem they had better solve.