The 2010 election was roughly 52% vs 45%. 45 million vs 39 million votes.
2006 was 52% vs 44%. 42 million vs 35 million votes.
Both 2010 and 2006 saw power shift from the majority party to the minority party in the house.
What kind of % vote difference and roughly how many votes would it take for the democrats to recapture the house, which I believe is 25 seats since they currently have 193 and need 218 to win a majority? In the 2018 off year, I’d assume about 75-90 million people will vote in house elections.
I’m guessing if the democrats get ~45 million votes and the GOP gets ~40 million votes, that will shift the house. However that doesn’t take into account gerrymandering and its role in making the house more red.
Does anyone have a rough idea what the minimum ratio or number of votes is for the democrats to gain 25 seats? 53% vs 46%? 45 vs 38 million votes?
There is no way to answer the question. It depends in large part on where the votes come from; geographical shifts can be as useless in the House races overall as they were for Hillary in the Electoral College (with her famous 3 million).
Nate Silver could probably give a good estimate and probably will at some point, but I’d estimate that a 3% win gets them a slim majority. 2% it would be very close, 1% probably means they fall short.
According to Nate the average swing is -7.5% from the President’s margin. That would leave the midterm about -9.5%. While I doubt that would happen I think it places the house about 50-50 and pick 'em. Any further decline that way and it’s Hello, Speaker Pelosi again.
I made an argument (here) based on the Cook partisan voting index that the 50/50 line is somewhere around D+3. That does not include data from the most recent election, but seems like a reasonable jumping off point…
Total votes for house races across the nation is a meaningless number. The house of representatives is not determined by a single election. It is 435 seperate elections.
No matter how wide a margin a candidate wins by, it won’t help anyone else in the other 434 districts.
To follow up, the Cook partisan voting index page at wikipedia has been updated to include 2016 election results.
The median district is in fact R+3. So a national vote of D+3 puts the median in play.
Another way to look at it is the mean district is D+0.46 while the median is R+3. So the democrats need to overcome a 3.46 point structural republican advantage to have a shot at the house.
Both ways of looking at ignore the particular candidates running in any given race and assume differences in results based on this tend to even out.
There are clear indications over the years that elections build in waves. It’s why we use generic ballot indices to such an extent. It gives analysts a means by which to determine which way the electorate is changing over the near term.
In short, elections - even congressional elections - don’t happen in a vacuum. If there’s a national trend toward one party or the other - typically against the president’s party during midterms - that trend will be evidenced in the vast majority of congressional districts. They don’t occur as independent events.
This is why elections like the recent special in Kansas - which the D lost by 7 but was in a R+20-something district - and the one in Georgia - which had been R+6 but the D candidate led the pack with 48% - are so important. It’s early yet but it gives an indication of how MUCH the anti-president’s party lean will be in November 2018.
Total votes for house races across the nation correctly predicts control of the house in 9 of the last 10 elections. Adjusting that extremely simple model by a 3 point(ish) republican structural advantage puts things at 10 out 10 correct predictions.
Total votes for house races across the nation is highly predictive. That is opposite of meaningless.
In 2012 the Democrats won the popular vote in the house, 48.8% to 47.6%. Despite that the gop held 234 seats vs 201 for the Democrats. So just winning. The popular vote nationally isn’t enough due to gerrymandering and urban vs rural divide issues.
So as others are saying, a 3% margin may be the minimum for the house to shift. Anything beyond that just increases the house margin.
On the subject of vote margins, in 2000 the gop got 47.6% of the vote, Democrats got 47.1%. this resulted in the gop having 221 seats, the dems having 212.
I know there was a lot of gerrymandering after 2010 and I know that not all districts are created equal with regard to population, but how did a 0.5% popular vote margin for the gop in 2000 result in 221 seats, but a -1.2% margin in 2012 resulted in 234 seats?
Was it all gerrymandering? Or is it population growth and shifts between districts also?
There’s a difference between honest redistricting because of population shifts, and gerrymandering – abut both can lead to either party’s advantage. Missouri, for example, lost one seat after the 2012 census. It got taken out of the St. Louis area.
You might feel that was gerrymandering to get rid of a seat that usually went Democratic. Or you might feel that because the St. Louis area had a stagnant population while other parts of the state were growing, and because Missouri’s rural districts were already geographically huge, it was the rational choice.
In either case the result was that a usually Democratic urban-suburban district was dissolved with its voters being divided into an always-Democratic urban district and an always-Republican rural district.
Or to put it yet another way: A massive landslide-to-landslide swing in one district would make a significant dent in the nationwide vote totals, but wouldn’t help the Democrats much, because it’d only be one seat… but massive landslide-to-landslide swings in just one district just don’t happen. It’s much more likely that a bunch of safe D districts become even safer, and a bunch of leaning-D districts become safe, and a bunch of safe R districts become leaning-R, and a whole lot of districts in the middle shift from barely-R to barely-D. The shift happens everywhere, including in the extreme districts where it mostly doesn’t matter, but where it does matter is in that big middle area.
Now, there will always be a handful of districts that buck the trend, one way or the other, for a variety of reasons: A major factory in that district just opened or closed and one party got the credit or blame, a candidate is particularly strong or weak, etc. So you will have some places where the swing is even stronger than it is nationwide, or where the swing doesn’t occur or is even reversed. But on average, you’ll have about the same number in both directions, and it mostly averages out.
I don’t think it would, actually. Each district is going to be on average less than .25% of the population. So a complete swing of the entire electorate couldn’t amount to more than a .5% swing in the nationwide results.
Political shifts, population shifts, and gerrymandering. FWIW I usually refer to the Washington Post’s analysis of post-2010 gerrymandering; it’s unsurprising they conclude that gerrymandering gave the Republicans a net advantage, but that net advantage was estimated at around 6-8 seats. That’s not nothing, but it’s also smaller than the GOP margin has been at any point from 2010 to present.
Political shifts & population shifts have gone hand in hand - Parts of the country that have typically been blue, major urban centers, are growing but attract a lot of “liberal” migration (and conversely, suburbs and exurbs attract more conservative migration, but to a less extreme degree, suburbs are often 55-65% GOP versus urban centers sometimes being as high as 90% Dem); meanwhile due to the structure of our geographical districting rural areas still retain an advantage even as they lose population. Political shifts have also seen an erosion of some traditional Democratic bases. In a lot of ways 2010 was the true “end” of the right-of-center and “conservative” Democrats, I think there’s like 2-3 of those guys still left in Congress. One of Minnesota’s Congressmen fits that mold, as does WV Senator Joe Manchin, but the reality is Democrats are just less viable in right-leaning districts now than they have been at any point in history.
You’ve also had a “political squeeze” that’s hit the “blue collar white” base that often used to lean Dem. Dems used to hold this group for economic reasons and it was often marshaled by unions, but the GOP and large corporations have successfully waged a 30 year war against unionization. Sans unions, the blue collar white vote is less concentrated and less organized. At the same time, the Dems have made decisions to put issues like “systemic racism” and “bathroom access for trans people” at the forefront in a lot of debates and the court of public opinion. This is kind of the larger story behind the Hillary Democrats versus the Sanders Democrats. Hillarycrats promote reasonable, left of center economics (I’m a conservative and a free marketer, so I don’t fully agree with them) that would have more benefit for most people than the current Republican economics (the GOP isn’t a free market party any longer, it’s basically a crony-capitalism party that just lies to its base and serves the 1%.) Hillarycrat economics isn’t super interesting, there’s no dramatic rallying cries in it because it represents incremental changes and reforms, similar to much of Obama’s economic policy. So for that reason the Hillarycrat’s playbook of stuff to “fire up the base” tends to be limited to social justice issues, and basically calling Republicans bigots. There’s good evidence this is a loser strategy for winning white votes because a lot of whites don’t buy that everyone who isn’t a Hillary voter is a racist, and they care more about economic pitches. Sanderscrats have basically been advocating for a focus more on a populist economic rhetoric, featuring (dangerous and wrong) accusations against big banks, the 1%, promises of free stuff and etc. Other than universal healthcare which I agree with, I think most of the Sanderscrat economic agenda is dangerous and stupid, but it’s undeniable it’s popular and that’s because you can sell it with flashy, simple platitudes. It’s hard to sell incrementalism in economics.
These political shifts are a big part of why the Dems have been losing State/Congressional elections throughout the Midwest–Ohio, Wisconsin, Michigan and Indiana are all fully controlled by the GOP. Minnesota has Republicans with a majority in both houses with a Dem Governor. Similarly Pennsylvania has legislative GOP control with a Dem Governor.
SUPPOSE the African-American turnout is much heavier in 20178 than it was for Hillary. Will that elect more Democrats to the House? Not necessarily. It might just mean that incumbent Democrats are re-elected by bigger margins, while all the Republicans are ALSO re-elected comfortably.
You can’t just get higher Democratic turnout- you need high Democratic turnout in potential swing districts.
I liked your post about Hillarycrats vs Sanderscrats. How would you describe the “social” / non-economic positions of the Sanderscrats? Sanders himself is thought of as generally more pro-gun than your average Democrat, but beyond that, I’m not sure how his supporters come down on most “social” issues. Your thoughts?