House forecast: Democrats will win 228 seats (and the House majority) while Republicans will win just 207 seats. A Democratic win of 204 seats and 263 seats is within the margin of error.
Senate forecast: Republicans will hold 52 seats (and maintain control of the Senate) next Congress while Democrats will hold just 48. Anything between Republicans holding 47 seats and 57 seats is within the margin of error.
Polls show that Democratic voters are revved up for November. There is one exception to that rule, however: Latinos. They seem far less enthusiastic than pretty much every other part of the Democratic base, and our forecast shows Democrats have a problem in Latino heavy districts.
I examined the five congressional districts in the country currently controlled by Republicans and where at least 50% of the citizen voting age population is Latino. These districts are not surprisingly in California, Florida and Texas.
In every single one of them, the Democrats are underperforming Hillary Clinton's numbers in the 2016 election.
California 21: Clinton took this district by 16 points. Democrat TJ Cox is forecasted to lose by 7 points.
Florida 25: Clinton lost the district by just 2 points. Democrat Mary Barzee Flores is forecasted to lose by 8 points.
Florida 26: Clinton won the district by 16 points. Democrat Debbie Mucarsel-Powell is forecasted to win by less a point.
Florida 27: Clinton won the district by an astounding 20 points. Donna Shalala is forecasted to win by less than 4 points.
Texas 23: Clinton won here by 3 points. Democrat Gina Ortiz Jones is forecasted to lose by 6 points.
All told, the Democratic candidates are underperforming Clinton by an average of 14 points in districts where at least 50% of the citizen voting age population is Latino.
Part of the reason Democrats are underperforming has to do with low turnout among Latinos. In a September poll of Texas' 23rd District by Siena College and The New York Times, Latinos were projected to make up 49% of 2018 voters. They were 65% of all registered voters in the district. An analysis by Politico of all the Siena College/New York Times polls indicates that this split is indicative of districts nationally.
It also seems likely that some Latinos who voted for Clinton in these districts are pulling the lever for Republican candidates in 2018. In a recent Mason-Dixon poll of Florida's 27th Congressional District, non-Cuban Latino and Cubans were 20 to 40 points more likely to vote Republican this year for Congress than they were two years ago for president.
Now, it should be noted that these districts have historically been at least more Republican than the 2016 presidential election would indicate. In all of them, the weighted average partisanship (which takes into account past state and local elections in addition to the presidential election results) is more Republican than the 2016 presidential vote.
The ability of the Republican candidates in these districts to continue that divergence (between voting patterns for 2016 president and the House) should at least be partially attributed to these candidates running towards the center.
Of the four incumbent Republicans running in these five districts, Rep. Will Hurd of Texas has the most conservative voting record. He's still more moderate than about 75% of the Republican caucus. Additionally, he's been an outspoken critic of Trump.
These moderate records have, at a minimum, probably quelled the anger of some Clinton voters to the extent that they don't feel the need to show up in November.
Now, it's not a guarantee that Democrats will come away empty-handed from majority Latino districts that are currently represented by a Republican. They'll probably win Florida's 26th and/or 27th district.
The fact that is far from a guarantee, however, speaks to the major problem Democrats seem to have with Latino voters heading into the 2018 midterms.