US mid-terms minus four days

Democrats remain likely to win the House, while Republicans are likely to hold the Senate at US mid-term elections on Wednesday (Australian time). Guest post by Adrian Beaumont.

Guest post by Adrian Beaumont, who joins us from time to time to provide commentary on elections internationally. Adrian’s work on electoral matters for The Conversation can be found here, and his own website is here.

The US mid-term elections will be held on November 6, with results coming in on Wednesday from 10am, Australian Eastern Daylight Time (AEDT). In the FiveThirtyEight poll aggregate, Democrats lead by 8.3 points in the race for Congress, slightly up from 8.2 points last week. In the FiveThirtyEight Classic Model, Democrats have an 85% chance to win the House, unchanged on last week, but down from 86% on October 30. Strong fundraising for Democrats has affected the classic model’s fundamentals calculation, and the polls-only “Lite” Model gives Democrats a 77% chance to win the House, unchanged since last week.

Democrats are rated just a 16% chance to win the Senate in the FiveThirtyEight Classic Model, down from a 17% chance last week. If Democrats win the five close states – Florida, Indiana, Missouri, Arizona and Nevada – where FiveThirtyEight has them slightly favoured, they would still need one of Texas, Tennessee or North Dakota to win Senate control, and those are currently likely Republican. There appears to be some narrowing in Republican Ted Cruz’s lead over Democrat Beto O’Rourke in Texas, but no poll has O’Rourke winning. As I said last week, the best chance for Republicans to retain the House or Democrats to win the Senate is a systematic polling error.

Trump’s ratings in the FiveThirtyEight aggregate are currently 42.1% approve, 52.9% disapprove, for a net approval of -10.8 points, down from -10.1 points last week. Trump has closed out the mid-term campaign with vitriolic anti-immigrant rhetoric. This rhetoric risks further alienating educated people, and costing the Republicans in House districts with high levels of educational attainment. However, the close Senate contests this year are in states without high levels of education. In my opinion, Trump would have done better to close out by focusing on the economy – unemployment is under 4%, GDP growth is solid, wages are growing in real terms and inflation is low. Stock markets improved this week.

Early voting in 27 states has already passed the total number of early votes recorded at the 2014 mid-term elections. In Texas, early voting has surpassed the total 2014 turnout! As early voting has gone on, there is clear evidence that young people are voting late in the early voting period. We can already say that 2018 will not be a normal low-turnout mid-term election. Democrats could overperform their polls if they vote at a greater rate than expected, relative to Republicans.

Poll closing times

All times listed here are Wednesday November 7 AEDT. Some states straddle two time zones. In this case, networks will not call a state, and exit polls will not be officially released, until all polls in the state are closed. I will concentrate on poll closing times for the key Senate races below. Source: The Green Papers.

10am: Indiana Senate, eastern zone. Most of Indiana is in this zone, while a small part closes at 11am.

11am: Florida Senate, eastern, and Virginia. The part of Florida that closes at 11am is relatively Democratic-friendly. The deeply conservative “panhandle” closes at noon, and will assist Republicans. Several House races in both Virginia and Florida are contested.

11:30am: West Virginia Senate, where Democrat Joe Manchin is expected to win in a state that vote for Trump by a crushing 42 points.

12noon: Illinois, New Jersey, Pennsylvania, Michigan (eastern) and Tennessee, Missouri and Texas (eastern) Senate. New Jersey, Illinois, Texas, Michigan and Pennsylvania have many contested House seats. Most of Michigan and Texas are in the eastern zone. Republicans are a long shot in New Jersey Senate, and Democrats are a long shot in Tennessee and Texas Senate. Missouri Senate is expected to be close.

1pm: New York, Minnesota, Colorado, Wisconsin and Arizona Senate; many House races are contested in these states, and the Arizona Senate is close.

2pm: Nevada and North Dakota (eastern) Senate. Republicans are favoured to gain North Dakota, while Nevada is expected to be close. Polls in the trailing part of North Dakota close at 3pm.

3pm: California and Washington State, where many House seats are contested.

Vote counting will finish late Wednesday night AEDT in California. Many states also have much post-election counting, which can take up to four weeks in California. This additional counting usually favours the Democrats, so a Republican election-night lead of 52% to 48% in a Californian district is not safe.

Author: William Bowe

William Bowe is a Perth-based election analyst and occasional teacher of political science. His blog, The Poll Bludger, has existed in one form or another since 2004, and is one of the most heavily trafficked websites on Australian politics.

31 comments on “US mid-terms minus four days”

  1. Me too Ven, but I don’t really expect the Senate.

    The polls in 2016 weren’t quite as far off as people make out, but I’m hoping they are now overcompensating, and the numbers for Trump are even worse than they look. As happened at the last UK election when the shy Tories didn’t show up.

  2. AB,

    I’m hoping there is a reason for the Trump strategy you describe, that they have given up the House and are not entirely confident about the Senate.

  3. AB,

    I’m hoping there is a reason for the Trump strategy you describe, that they have given up the House and are not entirely confident about the Senate.

  4. Just about all the polling points to the Democrats taking the House, and the GOP holding the Senate. In essence, the battle for the House is being fought in suburban areas of cities, areas that have traditionally leaned Republican, but who are apparently turned off by Trump’s focus on the ugly side of the Right and Trump’s general crassness. The Senate, on the other hand, is being fought in empty “Red” states where Democrats often won unexpectedly in Obama’s re-election in 2012 (think Indiana, Missouri, Montana and West Virginia), but who voted for Trump n big numbers in 2016.

    Consequently, what we are seeing this year is really two separate elections in different parts of the country. The good news for the Democrats is that if they can hold the line this year, they will face a much more friendly Senate map in 2020, as Republican Senators elected in 2014 will be up for re-election.

    The results next week will, of course, come down to turn-out, and reports of a surge in early voting this year bode well for the much-trumpeted “Blue Wave”. Indeed, turn-out is looking more like a Presidential year than a mid-term, which augers well for the Blue team. While Heidkamp looks gone in North Dakota, Tester in Montana and Manchin in West Virginia look pretty certain to get back, while McCaskill in Missouri, Nelson in Florida, and Donnelly in Indiana look better than evens to get back back. If the Dems can also pick up Nevada and Arizona they will draw level, but will need to rely on an upset in Texas or Tennessee to get a majority. Still, if turn-out is way up, then this is not completely out of the question.

    Even if the Dems do take the Senate, though, I don’t think impeachment is necessarily more likely. Conviction requires a two thirds majority of the Senate, and in any event, most of those Red State Democrats won’t vote to impeach Trump. What a Democratic Senate might do, however, is stem the flow of conservative judge appointments.

  5. Hugoaugogo, what’s necessary for a massive “blue wave” is not high turnout per se, but differential turnout. That is, if Democrats vote as if it’s a Presidential election, and Republicans vote as if it’s a mid-term election, Democrats will win a landslide.

  6. Adrian, yes that’s true of course, and I did try to couch my language around turn-out to allow for that sort of nuance. But most measures of voter enthusiasm seem to indicate that Democrat voters are a good bit more motivated than Republicans this year, and the surge in early voting may well suggest that this will be borne out on Election Day (and beforehand). Statistically there are more Democrats than Republicans, but traditionally the Blue voters are much less reliable at actually turning up to vote. Could be that the groper-in-chief is providing a powerful motivation to Democrats. Turns out voting actually does matter after all.

  7. A significant part of the Republicans backing Kavanaugh, rather than find another conservative judge, was to create a partisan issue to drive republican turnout. Trump`s campaign against the Migrant Caravan is also mainly for the purpose of increasing republican turnout, as is his increase in anti-socialist rhetoric.

    The ability for increased turnout to elect Democrats varies according to the proportion of the population who can but don`t usually vote but leant towards the Democrats. In small rural and mostly white heavily Republican dominated states, this is probably relatively low. In diverse state with significant urban populations but historically low turnout, the Democrats have a greater opportunity to turnout more of the vote. Therefore Texas is one of the more interesting races.

    Also, if the Democrats got a Alabama Special Election 2017 type result in Mississippi, they could make 2 gains as there is a special election there. That probably won`t happen but the outside change is there.

  8. Tom

    That second special Senate election in Mississippi is bizarre. It is essentially a “jungle primary” or a European-style “first round” election. If no candidate gets over 50%, there will be a run-off later. It would seem unlikely for the Democrats to win either but you never know.

    It must be strange (and confusing for the public) to be in a state when there are two concurrent Senate elections. I gather that in Arizona there is not a special election to replace John McCain’s seat (which has been filled by a Governor appointee – a former Senator who won’t run again) because there wasn’t time to have primaries etc. after he died.

  9. mick

    In the unlikely event that happens, and if the Democrats look competitive, I think the economy of Mississippi (one of the poorest states) would experience a “mini-boom” with politicians, campaign workers and media people from all over the USA descending en masse for three weeks before the November 27th “run-off” election.

    It would actually be fascinating if it were a competitive race and one to decide Senate control. It would highlight the many problems of areas and states “left behind”. Healthcare would have to feature prominently as I think Mississippi’s stats are just about the worst along with Alabama and Arkansas from memory.


  10. Hugoaugogo says:
    Saturday, November 3, 2018 at 9:41 pm
    The Senate, on the other hand, is being fought in empty “Red” states where Democrats often won unexpectedly in Obama’s re-election in 2012 (think Indiana, Missouri, Montana and West Virginia), but who voted for Trump n big numbers in 2016.

    Indiana cannot be classified as empty “Red” state. Trump carried it with huge margin because Pence was VP pick. In 2008 Obama carried Indiana. Hopefully Donnelly wins it but current aggregate poll has Repug candidate in front.


  11. Hugoaugogo says:
    Saturday, November 3, 2018 at 10:20 pm
    Statistically there are more Democrats than Republicans, but traditionally the Blue voters are much less reliable at actually turning up to vote.

    However there about 20 % (maybe more) other voters who are classified as Independents. IMO, 70 % of those are Republicans or Republican leaning voters who although like Republican party policies do not want to be registered as Republicans because Republican party are taken over by nutters. Now a days, We see them on regular basis on CNN & MSNBC criticising Republican party or the suburban educated decent to high earners. IMO, They are the key to this Mid term election because democrat voters are not committed voters like Republicans

  12. I’m not sure I’ve ever paid any attention to the midterms before except perhaps a casual reflection on the scores in the aftermath, but this one has been hilarious.

    Some definite takeaways for home here in Australia, firstly the guys in the US are getting a lot better at recognising the asymmetry between the two parties and rejecting the lazy pox on both houses.

    But it is huge fun to listen to the crooked media guys and other Democrat leaning guys talk about ‘turn out’ in such terrible fear. Also wonderful to hear Nate talk about his models worried not about just error in one direction but worried about them all. Love model talk.

  13. Ven – yes Obama did carry Indiana in 2008, but he was the only Democrat to do so since the LBJ landslide of 1964 (and Johnson himself was the first Dem to win it since FDR in 1936), and Obama only won by less than 1% in the teeth of the worst economy since the Great Depression. I think we can safely categorise as a Red State, which is why Donnelly is running how supportive he has been of Trump and his policies.

  14. “The guy Jon Ralston …”

    Nate Silver said something like ‘do not waste any of your life or your time listening to anyone say anything about early voting and what it might mean, except Jon Ralston’. There may have been a second exception but I don’t remember that bit.

  15. There appears to be a tightening in the polls which does not augur well for the Dems’ Senate chances. Chances of a sweeping House win also seem to be fading.

  16. I went to 538 to see confirmation of the “tightening” seemingly being referred to by media

    The latest poll referred to is Harris Interactive showing a margin in favour of the Democrats at 11%, out from 6% when they last reported

    There were others showing the Democrats up by 6%, being a tightening but now comes the Harris Interactive out to the referred 11% and one other at 10%

    So is this a “tightening”

    We will know soon enough

    The overall summary of 538 continues to show a margin to the Democrats at 8% and at 50 to 42 (decimal points about even)

    To me this election and any verifying of the Trump regime is a commentary on the American Nation (noting again the commentary of Stiglitz but also noting those at Trump’s rallies which is disturbing)

  17. This has been my favourite “day before” election piece that 538 does. Use it if you’re going to obsessively lap up each tidbit of results as they come in, to get the earliest sense possible of how the US midterm results are going to eventually land.

    https://fivethirtyeight.com/features/2018-election-polls-close/

    NB: set the “world clock” on your phone to Washington DC time. That corresponds to “EST” as 538 states it. That way your timing will be right for where you are.

  18. This is from the 538 article I linked just above. 9pm EST in the US will be 1:00pm EST tomorrow in Australia. Look out for calls on control of the US House about then.

    “9 p.m.

    Polls close in: Arizona, Colorado, Louisiana, the rest of Kansas, the rest of Michigan, Minnesota, Nebraska, New Mexico, New York, the rest of North Dakota, the rest of South Dakota, the rest of Texas, Wisconsin, Wyoming

    My guess is that this is the hour when we’ll know if Democrats will win the House (or at the very least, if they’re on track to do so).”

    https://fivethirtyeight.com/features/2018-election-polls-close/

  19. Nate Silver does not rule out any possibility. The likelihood is that the Dems will win the House by 30+ seats but Repug gerrymanders complicate the issue.

    A Democrat Senate win is possible but highly unlikely.

  20. Input from the latest polls to the 538 index on the Senate & the House showed a slight but noticeable movement toward the Democrats

  21. Ok an actual psephological question here. (Given a lot is partisan policy commentary, which I am sympathetic to). How do you get meaningful polls with non compulsory voting in a system as fragmented and complex as the US? They seem to get ok ones but I struggle to imagine how given all the variables.

Leave a Reply

Your email address will not be published. Required fields are marked *