As the evidence mounts that the Democratic Party’s formerly inevitable crushing victory in the upcoming midterm elections may be less than crushing—perhaps even less than a victory—liberals have zeroed in on the real problem: conservatives who’ve noticed.

Over the weekend, CBS News/YouGov released a poll that no honest partisan Democrat would describe as anything other than a letdown. According to the network’s “battleground tracker,” the most likely outcome of the 2018 midterm elections, were they held today, would be a net Democratic gain of just nine seats. That would flip the House, but with a majority of just 219 to the GOP’s 216 members. That’s a far cry from the 35- to 50-seat gain many were predicting for the “out-party” not that long ago. The Democratic Party’s declining fortunes is measurable across a variety of surveys by different pollsters. The Real Clear Politics average of “generic ballot polling,” which showed a Democratic advantage of over nine points as recently as late February, now shows a Democratic advantage of just over three points.

What’s more, the polling seems to be reinforced by the outcome of some recent special elections. Whereas Democrats once out-performed the polls in both special elections and in last November’s off-year races, Democratic candidates have recently been faring about as well as polls indicate they should. “In the median special election since April 1,” CNN analyst and former 538 elections guru Harry Enten wrote, “Democrats have been outperforming the partisan baseline by only about half a point.”

Along with Donald Trump’s improving job approval rating, these troubling signs have not gone unnoticed by elections analysts and liberal pundits. As FiveThirtyEight’s analysts noted on a recent podcast entitled “Things Are Looking Worse For The Democrats,” despite the uncertainty inherent in polling, it’s clear that the political environment is more favorable for the GOP today than it was yesterday. As early as late January, Democratic strategist Ed Kilgore warned his fellow liberals not to dismiss the Democratic Party’s declining prospects as “noise,” even though it might be just that. Though they continue to enjoy big opportunities with some core Democratic constituencies this fall, Democracy Corps pollsters cautioned in late April that Democratic “momentum has stalled.”

Over the weekend, I joined John Podhoretz in noting the extent to which Democratic fortunes have deteriorated over the course of the last four months. What followed was a cavalcade of excuses, indignation, and moral preening as a liberal firing squad was organized to shoot the messengers.

“Being mad at libs because they are winning a poll that’s probably statistically insignificant is a great way to excuse yourself for being part of a movement endorsing lawlessness,” wrote USA Today contributor Jason Sattler. “I dunno kinda seems like if you’re writing hot takes of any kind off of one poll in early June showing the House to be a toss-up 5 months from now, you really need to find better things to write about?” Nate Silver snarked. “Based on state-by-state analysis of legislative tactics & skewed outcomes, GOP has engineered a 3-4% advantage in the national vote,” Princeton University data scientist Sam Wang told Podhoretz. New York Times analyst Nate Cohn echoed this sentiment; due to the GOP’s “structural advantage” in the House, he said, the fact that the GOP’s modest 17-seat majority is even in jeopardy at all would suffice for a Democratic wave election.

It seems, however, that liberal political observers have internalized the data suggesting that Democratic prospects in November have soured. They just don’t want to hear about it from unacceptable company. Their objections might have to do with the fact that conservatives aren’t merely neutral observers watching Democratic prospects decay in a judgment-free vacuum. Conservatives have some theories that explain why this is happening to the left, and liberals don’t like that at all.

There are, of course, some relatively banal reasons that might explain why Trump and the GOP are up and Democrats are down: The economy is strong and improving and the United States is not presently engaged in any high-casualty foreign wars. But those conditions prevailed when Trump’s job approval was mired in the mid-30s and Democrats had a double-digit advantage in the generic ballot. Moreover, Democratic enthusiasm and engagement in the political process have not fluctuated much in that period, but Republican enthusiasm and engagement most certainly have. When a significant number of Republicans are disenchanted with Trump or fractured by GOP infighting, that is when Donald Trump’s job approval takes a dip. Likewise, when Republicans “come home” to Trump, his job approval ticks up along with the GOP’s prospects for November.

So if Democrats are buoyed when depressive, marginal Republican voters are discouraged, then the Democratic objective should be to discourage them. But instead, Democrats have convinced themselves that they cannot spend 2018 hammering Donald Trump on issues related to character and comportment. They need to advance the progressive cause by advocating their preferred policies—ranging from the imposition of a single-payer health-care system on the country to “free” college to compulsory unionization. What’s more, Democrats have convinced themselves that they need to make the case against not just Trump but the Republican Party and its legislative achievements. They do this by attacking GOP accomplishments such as like judicial appointments and the reformation of the corporate and individual tax code. All the while, Democrats have reminded marginal Republicans why they were Republicans in the first place. As long as Donald Trump manages to avoid stepping on any rhetorical landmines, that translates to increased enthusiasm among Republican voters.

It is perfectly reasonable to expect Democrats to react defensively to the notion that Democratic prospects for November are undermined when the party elaborates on what it actually believes. If you thought that the progressive program was as popular as polls suggest, despite the decimation of the Democratic Party over the course of the Obama years, you, too, would take maximum umbrage with that assertion. But if Democrats are as determined to avoid self-criticism today as they were in the Obama era, they might find that their electoral prospects remain similarly moribund.

I continue to believe that 2018 will be a good year for Democrats, just because it seems unlikely that Trump’s first midterm will join only 1934, 1962, and 2002 as midterms in the modern era that redounded to the president’s benefit. And yet, despite all his transgressions against civic comity, progressive orthodoxy, and common decency, the thought that Trump may not endure the repudiation they believe he is owed is psychologically traumatic to Democrats. If they do have a reasonably “bad night” in November, they would be well served to come to terms with the fact that an ideologically homogenous minority is still a minority.