It’s hard to identify the precise moment that President Trump’s reelection bid began. Was it in February, when he named a campaign manager? Was it in January 2017 — the day of his inauguration, in fact — when he filed the requisite paperwork? Was it the first day of 2017, when his campaign first spent money on the 2018 election cycle? Or was it Dec. 1, 2016, less than a month after his victory, when he held his first post-election rally?

Trump has been talking about his reelection since before his inauguration, making reference to serving for eight years during a news conference in New York during the transition period. He is reportedly ready to turn his attention to that contest more fully, now that the 2018 midterm election is over.

By at least one metric, though, the 2018 election should give him pause about the prospect of serving two full terms.

A colleague raised an interesting question: What does the House vote Tuesday night — the only contest in which nearly every American was asked to weigh in — tell us about the 2020 election? If we pit the votes for Democratic candidates in those races against the votes for the Republicans, where do we land?

The short answer is this: Democrats received about 4.4 million more votes than Republicans on Tuesday, according to the most recent tallies from the Associated Press. That’s while votes are still being counted, including in Democrat-heavy California.

Unlike in 2016, though, those votes were also distributed in a way that helps the Democrats. That margin translates to a 290-to-248 electoral vote edge for the Democrat.

UNADJUSTED ADJUSTED
AK
                   
ME
           
WI
     
VT
NH
 
WA
ID
MT
ND
MN
IL
MI
 
NY
MA
 
 
OR
NV
WY
SD
IA
IN
OH
PA
NJ
CT
RI
 
CA
UT
CO
NE
MO
KY
WV
VA
MD
DE
 
   
AZ
NM
KS
AR
TN
NC
SC
DC
   
       
OK
LA
MS
AL
GA
     
HI
     
TX
       
FL
   

That margin comes as a result of Democrats flipping four states that voted for Trump in 2016: Iowa, Michigan, North Carolina and Pennsylvania.

Those may not be the four states you’d expect, given that Trump’s 2016 victory was a result of close votes in Michigan, Pennsylvania and Wisconsin, not Iowa. He won Iowa by more than nine points (more than his margin in Texas) and North Carolina by four points. The House results in those states Tuesday, though, showed Democrats winning by four points in Iowa and two in North Carolina.

Both Iowa and North Carolina saw consistent shifts to the Democratic candidate in House races Tuesday, though — in Iowa by wide margins. (Shown below as larger blue boxes.)


(Philip Bump/The Washington Post)

But the data powering our assessment of 2020 is not all that great. Notice that we said the House contests involved votes from nearly every American. In several states, candidates were reelected without opponents. That includes Florida, where the vote total for the GOP was narrow enough that uncontested Democratic reelections could have made the difference.

(There are also places where multiple candidates from the same party faced off, boosting the vote totals for those parties. But those were in California, Louisiana and Washington, where the results weren’t really in question, so we skipped them.)

To offset the uncontested races, we created an adjusted version of the data. To estimate the missing vote totals, we assigned each race a number of Democratic and Republican votes equivalent to the average across the state from contested races in which the same party ended up in control. So in Florida, a Democrat elected without an opponent was assigned vote tallies equivalent to the average of what was seen in other Florida House districts that elected Democrats.

The result? Not much different. Florida moved a bit to the Democrats, but the Republican candidate (presumably Trump) would still win. Wisconsin moves from a five-point Republican advantage to a 0.4-point one — less than the margin Trump had in 2016. North Carolina moved to the Republicans, but, again, didn’t flip.

There are innumerable caveats here, drifting down on us like snowflakes. House races involve hundreds of candidates and hundreds of policy priorities. Presidential races tend to operate on a much narrower scale. What’s more, the electorate for a midterm election, particularly this midterm election, may not look much like what we will see in 2020.

And, that year, Trump is expected to be on the ballot, which might have an effect similar to what we saw between 2010 and 2012: a surge in the latter year among supporters of the president who were less inspired to vote in 2010. Then, of course, there’s the other candidate, whoever he or she is. That person might spur his or her own surge in turnout.

They’d better. After all, Trump has a one- to two-year head start on campaigning.