Democrats are right to be worried by polls showing their falling midterm margins. They got lost following this map two years ago. The reason for their loss then, and their legitimate worry now, is that President Trump won his votes to maximum electoral vote impact and that impact closely replicates today’s congressional map.

Polls clearly show Tuesday’s races tightening. Rasmussen’s last generic ballot poll before the election had Republicans narrowly pulling ahead of Democrats 46-45 percent — erasing last week’s three-percent deficit. While one poll does not a trend make, it is clear others also show a shrinking gap.

Further, President Trump’s ratings are rising. Just three days ago (11/2/18), Rasmussen’s daily tracking poll showed his approval rating at +4 percent (51 percent approval to 47 percent disapproval). (If that does not seem overwhelming, consider he won the White House with 46 percent of the popular vote, while Clinton received 48 percent: That means Trump is +5 percent over his November 2016 total, and his disapproval at -1 percent below Clinton’s.) Within those figures, his strong approval rating was 37 percent and his strong disapproval rating was 40 percent. As a proxy for intensity, -3 percent means the gap is virtually nonexistent.

For all practical purposes, Democrats’ hopes of taking the Senate are gone. Now their hope shifts to minimizing losses. Similarly, Democrats’ visions of a large House majority are now down to just winning the House — even that is no longer a certainty. The blue wave threatens to become a ripple.

Democrats have seen expectations dashed before. Two years ago, their presidential candidate won the popular vote by over two percentage points, yet lost the election in an electoral vote landslide. This dichotomy is more than an unpleasant memory for Democrats; it is a warning because 2016’s electoral vote map mirrors today’s congressional battleground.

Clinton won the popular vote by 2.9 million votes. Averaged across the fifty states, Hillary could have won each by over 57,000 votes. Democrats’ problem is that they were not; instead, Clinton racked up a massive 4.3 million vote win in California, but lost the rest of America by 1.4 million votes. This discrepancy in vote placement translated into Trump winning 306 electoral votes.

Far from being ancient history, this is Democrats’ current adversity with the Trump political map. Trump’s 306 electoral votes extended over 30 states. That equates to 60 Senate seats and 246 House ones. Of course, Trump did not win all those states’ congressional districts, but Republicans still won 235 overall — well more than half of Congress’s two bodies.

Trump’s ability to win his votes to maximum advantage is a multiple warning to Democrats.

Even if Rasmussen’s latest generic poll is wrong, simply being ahead in nationwide generic polls will not do for Democrats. Going into Election Day 2016, Democrats believed themselves ahead by secure margins. It was not enough. More to the point, even being ahead by two percentage points in the actual popular vote nationwide was not enough.

Democrats need to be ahead in their specific races. Racking up big protest votes in deep blue areas will not help them.

Another concern coming from 2016 should be just how accurate the national generic ballot polling is. Again, it showed them winning then, but was clearly wrong — the race was closer than most thought. Democrats must now worry if today’s polling is no better.

Democrats’ further concern must be attrition in their support. It is one thing to answer a poll question; it is another to actually go to the poll. While this would seem to be a worry for both sides, it is a disproportionate one for Democrats.

In the last three midterms — 2014, 2010, and 2006 — exit polling showed liberals falling as a percentage of the electorate from the previous presidential election (2012, 2008, and 2004). Of course, liberals are core Democratic supporters.

Over that period, liberals averaged a 1.7-point drop as a percentage of the voting population. Conversely, conservatives, core Republican supporters, increased as a portion of the electorate in two of those three midterms. They have averaged a 2.7 percent increase in the midterm voting population from the previous presidential election.

Taken together, these caveats about seemingly picayune percentages add up to bigger worries than Democrats would care to admit. As 2016 showed, they are neither hypothetical nor inconsequential.

As big as 2016’s Democrat disappointment turned out to be, 2018’s could be too. Polling is already tightening. As in 2016, polling may again be overstating Democrat strength. Even if accurate nationally, will that translate into the numerous separate races Democrats are seeking to win — again, it did not in 2016.

Even if accurate and favorably spread across the electoral map, will Democrats be able to survive the attrition in their liberal base that has occurred in the last three midterm elections? History shows Republicans do not have this concern with their conservative base.

In 2016, President Trump won his minority of votes to maximum advantage; Hillary Clinton won her majority of votes to minimum advantage. As Democrats approach the midterms they have eagerly anticipated for two years, there must be growing apprehension that disappointment could repeat itself.

J.T. Young served under President George W. Bush as the director of communications in the Office of Management and Budget and as Deputy Assistant Secretary in legislative affairs for tax and budget at the Treasury Department. He served as a congressional staffer from 1987-2000.