Election Update: The Romney-Clinton districts are overrated. Obama-Trump districts are underrated.

This year's mid-term did not really present the "model wars" we experienced in 2014 or 2016 – heated discussions between different election forecasters, whose forecasts sometimes yielded very different results. Instead, the different electoral models and forecasts have widely told the same story, one where the Democrats are strong but not some of the favorites to win the House, and the same goes for Republicans in the Senate.

But there was some The differences of opinion, and one of them concerns the districts most likely to be held by Democrats in the House. According to one theory, mid-term reviews are essentially a referendum on President Trump. This theory postulates that Democrats are likely to cope very well in the districts – often mixed, urban or suburban – that have recently turned into democrats, such as the 13 districts that voted for Mitt Romney in 2012 and Hillary Clinton in 2016. Conversely, the Democrats The theory would say that it is difficult to fight for gains in the predominantly rural and white districts that have moved the most recently to Republicans – like the 21 districts that voted for Barack Obama. in 2012 but Trump in 2016.

A contrary theory The mid-term is mainly about the return to the average, and about the voters trying to balance the power between the two political parties. According to this theory, districts that recently switched to Trump but were purple enough to vote for Obama in 2012 could be hot spots for Republicans. This is because they tend to be "swinging" (or elastic) districts and do not necessarily have much long-term loyalty to the Republican Party. In addition, to the extent that Trump won these states and districts on the basis of marginal voters with late decisions – which was the case in the Midwest, which is home to many Obama-Trump districts – these voters might not be as reliable. Republicans in 2018. Sometimes the most recent voters to join your coalition are the first to return immediately.

These theories do not exclude each other completely. The playing field of the House is quite broad, with Democrats having opportunities in all types of districts controlled by Republicans; that's what makes their chance of getting a more robust majority. Nevertheless, the Romney-Clinton districts have received a lot of attention and even speak of a "suburban tsunami"; the Obama-Trump quarters, less.

But if you look at the polls, the problem is that the Democrats are doing just as well in the Obama-Trump districts. Probably a little better, actually. First, here are the current adjusted FiveThirtyEight survey averages in the Romney-Clinton districts, all of which have at least one survey:

On average, Democrats have a 1.5 percentage point lead in polls in these districts. That's pretty good considering that most of them have Republican elected officials, among whom only the representative of Will Hurd in Texas 23 has a decisive lead in the polls. But it's not spectacular either. Clinton won these districts with an average of 4.7 percentage points – more than Democratic congressional candidates (even though Obama lost them by 7.8 points on average).

And here are the Obama-Trump districts … and the Obama-Trump states! Trump won six states that Obama won in 2012, five of which – Iowa is the exception – are running a race in the US Senate this year. Clinton did not win any state of Romney.[/footnotes].

As you can see in the table, four Obama-Trump congressional districts do not have an investigation, perhaps reflecting the less attention given to the media this year. In these districts, for example, I used our CANTOR system, which estimates the results of polls in these districts, if any, based on surveys in similar districts. (This is not a perfect compromise, but three of the four districts rate "Solid D" according to FiveThirtyEight and Cook's political report, and they would likely show important democratic leads if they had been questioned, which would distort sample, exclude them.)

This is an eclectic mix of neighborhoods, many of which are slam dunks for Democratic candidates in place and others where Republicans are favorites. But on average, Democrats are leading with 5.5 percentage points in the polls (counting CANTOR districts for which we used imputed polls). That's roughly equivalent to Obama's 6.1 point victory in those districts in 2012 and is much better than Clinton's loss to 7.1 points. Excluding the districts of CANTOR, the average electoral advance of Democrats is 2.3 points in these districts.

Meanwhile, Democrats are doing extremely well in the five Obama-Trump states with Senate races this year. On average, their candidates are leading with 12.5 percentage points, and only Bill Nelson, of Florida, is in real danger of losing. These candidates are all incumbents, and you would expect them to beat the partisan pole of their districts, but Obama was also a titular in 2012, and he earned those states by an average of 5.1 percentage points .

So for my money, the democratic performance is a bit better globally in the Obama-Trump districts than in the Clinton-Romney districts. You can credibly say that the problem is about the same if you take into account that the number of Democratic incumbents is higher in the first group than in the last group. But at a minimum, the polls did not really match the media story about the good results of Democrats.

For every Romney-Clinton district where Democrats excel in polls, such as Kansas 3 or Virginia 10, there are others – Illinois 6, for example, or several districts in California – where they have an average number. And although there have been some Obama-Trump districts where the number of Democrats has been low lately – especially Minnesota 8, in the Iron Range of Minnesota – these are more exceptions than the rule. More typical is Maine 2, a trumpy, swingy, white, secular, rural district where Democrat Jared Golden has taken a tie or perhaps a slight lead in polls against Republican President Bruce Poliquin.

Another way to address this issue is to see the correlation between the polls and the results of the presidential elections of 2016 and 2012, respectively. And that's what I did in the final table, below. I have established correlation coefficients between the average FiveThirtyEight polls in each poll in the House and Senate poll, as well as their results in the previous two presidential elections. Correlations are weighted according to the number of surveys conducted in states and districts. As a result, places with newer and more robust surveys have more say in the calculation.

This year's congressional race polls are actually more correlated with 2012 presidential results (correlation coefficient of 0.74) to those of 2016 (0.61). The effect is somewhat reduced if you take into account the occupation status of the districts. But based on the polls, the map looks at least as much as in 2012 as in 2016. Maybe the midterm elections are not all about Trump after all.

It is not a coincidence that this was also the trend in the various special elections in the country this year and last year: their results were more in correlation with 2012 than with 2016. For example, the worst result among the special elections for Democrats was Jon Ossoff in Georgia 6, the wealthy suburb of Atlanta who voted for Romney by far in 2012 but almost for Clinton in 2016 win the House – but if they meet the majority, do not be surprised if their route involves more Maine 2 and less Georgia 6.