Friday 25 November 2022

Victoria - The Place to Be

The Race:

Despite a number of media outlets publicly supporting the Coalition, the general consensus seems to be that the Andrews Labor government will be comfortable returned in the lower house. We will examine this by beginning with each seat's two-party preferred ALP margin (marked negative where held by the Coalition) and reducing this by 2.8% (i.e. the swing from the 2018 2PP results to the latest poll [paywalled]).

SEAT PARTY MARGIN PREDICTION
Albert Park  ALP 13.1% 10.3%
Ashwood  ALP 2.0% -0.8%
Bass  ALP -0.7% -3.5%
Bayswater  ALP -0.6% -3.4%
Bellarine  ALP 11.4% 8.6%
Benambra  LIB

Bendigo East  ALP 12.1% 9.3%
Bendigo West  ALP 18.6% 15.8%
Bentleigh  ALP 11.4% 8.6%
Berwick  LIB -1.3% -4.1%
Box Hill  ALP 3.1% 0.3%
Brighton  LIB -0.5% -3.3%
Broadmeadows  ALP 25.2% 22.4%
Brunswick  GRN

Bulleen  LIB -5.5% -8.3%
Bundoora  ALP 16.2% 13.4%
Carrum  ALP 12.0% 9.2%
Caulfield  LIB 0.0% -2.8%
Clarinda  ALP 14.9% 12.1%
Cranbourne  ALP 9.3% 6.5%
Croydon  LIB -1.0% -3.8%
Dandenong  ALP 23.1% 20.3%
Eildon  LIB -1.0% -3.8%
Eltham  ALP 8.8% 6.0%
Essendon  ALP 15.8% 13.0%
Eureka  ALP 9.6% 6.8%
Euroa  NAT -15.8% -18.6%
Evelyn  LIB -1.8% -4.6%
Footscray  ALP 27.6% 24.8%
Frankston  ALP 10.2% 7.4%
Geelong  ALP 10.3% 7.5%
Gippsland East  NAT -17.6% -20.4%
Gippsland South  NAT -14.2% -17.0%
Glen Waverley  LIB -0.9% -3.7%
Greenvale  ALP 22.0% 19.2%
Hastings  LIB 0.0% -2.8%
Hawthorn  ALP 0.6% -2.2%
Ivanhoe  ALP 12.3% 9.5%
Kalkallo  ALP 20.9% 18.1%
Kew  LIB -4.7% -7.5%
Kororoit  ALP 25.3% 22.5%
Lara  ALP 19.1% 16.3%
Laverton  ALP 23.4% 20.6%
Lowan  NAT -21.1% -23.9%
Macedon  ALP 13.4% 10.6%
Malvern  LIB -6.0% -8.8%
Melbourne  GRN

Melton  ALP 5.0% 2.2%
Mildura  IND

Mill Park  ALP 24.9% 22.1%
Monbulk  ALP 7.1% 4.3%
Mordialloc  ALP 13.4% 10.6%
Mornington  LIB -5.0% -7.8%
Morwell  IND

Mulgrave  ALP 15.8% 13.0%
Murray Plains  NAT -24.0% -26.8%
Narracan  LIB -10.0% -12.8%
Narre Warren North  ALP 10.4% 7.6%
Narre Warren South  ALP 10.4% 7.6%
Nepean  ALP 0.7% -2.1%
Niddrie  ALP 12.5% 9.7%
Northcote  ALP

Oakleigh  ALP 16.0% 13.2%
Ovens Valley  NAT -12.1% -14.9%
Pakenham  LIB 2.2% -0.6%
Pascoe Vale  ALP 22.3% 19.5%
Point Cook  ALP 12.8% 10.0%
Polwarth  LIB -2.0% -4.8%
Prahran  GRN

Preston  ALP

Richmond  ALP

Ringwood  ALP 3.2% 0.4%
Ripon  LIB 2.8% 0.0%
Rowville  LIB -5.5% -8.3%
Sandringham  LIB -0.4% -3.2%
Shepparton  IND

South Barwon  ALP 3.0% 0.2%
South-West Coast  LIB -3.2% -6.0%
St Albans  ALP 22.0% 19.2%
Sunbury  ALP 14.5% 11.7%
Sydenham  ALP 18.3% 15.5%
Tarneit  ALP 17.9% 15.1%
Thomastown  ALP 27.4% 24.6%
Warrandyte  LIB -3.8% -6.6%
Wendouree  ALP 11.0% 8.2%
Werribee  ALP

Williamstown  ALP 19.9% 17.1%
Yan Yean  ALP 16.9% 14.1%

On these numbers alone, Daniel Andrews has the numbers to re-form majority government.

(Note 1: Narracan is marked as a Liberal retain, but due to the unfortunate death of the Nationals candidate, this vote will not be held in line with the other seats.)

(Note 2: Bass is assumed to flip to Liberal, but could flip to Nationals as both parties are running Coalition candidates. Liberals are preferred as the seat has historically been held by Liberals, not Nationals).

The Gaps:

This does not address the eleven seats where the 2PP race is not ALP v Coalition, and introduces one tossup: Ripon.

A similar approach can be applied to the five ALP v Green races, but relying on primary votes (PV) for calculating the swing: (polled ALP PV/(polled ALP PV + polled GRN PV)) - (2018 ALP PV/(2018 ALP PV + 2018 GRN PV)) = -4.0%. This assumes that all votes for other parties fall along similar lines as the ALP:GRN PV, which may not hold true at all. However, if it did the results would show:

SEAT PARTY MARGIN PREDICTION
Brunswick  GRN -2.0% -6.0%
Melbourne  GRN -1.7% -5.7%
Northcote  ALP 1.7% -2.3%
Preston  ALP 21.3% 17.3%
Richmond  ALP 5.8% 1.8%

And calculating a -2.8% swing in primary votes LIB v Green (Liberal > 0; Grn < 0) we can also resolve Prahran:

SEAT PARTY MARGIN PREDICTION
Prahran  GRN -8.2% -10.4%

That leaves all currently Green seats their verdant disposition, and suggest Northcote will flip. I personally doubt this on a gut level, but the system is there to give better guidance to personal whims so I'll keep this as a highly speculative call. That's this election's wild call.

Benambra has been some marine shade of conservative since 1877. That an independent with a 16% primary vote came second in this race shows how far Labor and the Greens lag here. Jacqui Hawkins is running as an independent again, but I don't see any reason to think she'll be more successful this time around.

LIB

Mildura was won by a margin of 253 votes after preferences in 2018, so the margin is razor thin (0.0%). General polling across the state is meaningless on a case-by-case basis, but I'll suggest that the incumbent Ali Cupper now has the benefit or name recognition and is well placed to hold on for a second term.

IND

Morwell's independent is retiring, so reverting to the 2PP data the seat is considered to have a 4.0% margin by Labor which allows it to stay red even after a 2.8% swing.

ALP

Ripon is currently held by a Liberal Louise Staley, but following redistribution is considered marginally Labor at 2.8%--the exact size of the swing against Labor bringing this down to a pure 50-50 split. I'm going to assume that the voters brought into Ripon will be a little more inclined to vote for the Coalition than in their previous seats due to the incumbency of Ms Staley, and will keep this one blue.

LIB

Shepparton was a NAT stronghold, but has been held by an IND since 2014. With a 5.3% margin it's not unassailable, but I'm still tipping for an IND in a seat where support for the Coalition has marginally decreased but where the ALP is not a credible chance.

IND

Werribee is so deeply red that the Coalition didn't even get a second place. Labor still has a comfortable 9.1% lead on last election's IND, so probably a safe bet for the Andrews Government.

ALP

The Conclusion:

All tallies up this amounts to a prediction of 50 Labor seats, 32 Coalition seats and 6 cross-bench MPs, giving a reduced, but clear majority to the returning government. The seats to watch as most difficult to predict will be Northcote, Mildura and Ripon.

SEAT PARTY MARGIN
Albert Park  ALP ALP
Ashwood  ALP LIB
Bass  ALP LIB
Bayswater  ALP LIB
Bellarine  ALP ALP
Benambra  LIB LIB
Bendigo East  ALP ALP
Bendigo West  ALP ALP
Bentleigh  ALP ALP
Berwick  LIB LIB
Box Hill  ALP ALP
Brighton  LIB LIB
Broadmeadows  ALP ALP
Brunswick  GRN GRN
Bulleen  LIB LIB
Bundoora  ALP ALP
Carrum  ALP ALP
Caulfield  LIB LIB
Clarinda  ALP ALP
Cranbourne  ALP ALP
Croydon  LIB LIB
Dandenong  ALP ALP
Eildon  LIB LIB
Eltham  ALP ALP
Essendon  ALP ALP
Eureka  ALP ALP
Euroa  NAT NAT
Evelyn  LIB LIB
Footscray  ALP ALP
Frankston  ALP ALP
Geelong  ALP ALP
Gippsland East  NAT NAT
Gippsland South  NAT NAT
Glen Waverley  LIB LIB
Greenvale  ALP ALP
Hastings  LIB LIB
Hawthorn  ALP LIB
Ivanhoe  ALP ALP
Kalkallo  ALP ALP
Kew  LIB LIB
Kororoit  ALP ALP
Lara  ALP ALP
Laverton  ALP ALP
Lowan  NAT NAT
Macedon  ALP ALP
Malvern  LIB LIB
Melbourne  GRN GRN
Melton  ALP ALP
Mildura  IND IND
Mill Park  ALP ALP
Monbulk  ALP ALP
Mordialloc  ALP ALP
Mornington  LIB LIB
Morwell  IND ALP
Mulgrave  ALP ALP
Murray Plains  NAT NAT
Narracan  LIB LIB
Narre Warren North  ALP ALP
Narre Warren South  ALP ALP
Nepean  ALP LIB
Niddrie  ALP ALP
Northcote  ALP GRN
Oakleigh  ALP ALP
Ovens Valley  NAT NAT
Pakenham  LIB LIB
Pascoe Vale  ALP ALP
Point Cook  ALP ALP
Polwarth  LIB LIB
Prahran  GRN GRN
Preston  ALP ALP
Richmond  ALP ALP
Ringwood  ALP ALP
Ripon  LIB LIB
Rowville  LIB LIB
Sandringham  LIB LIB
Shepparton  IND IND
South Barwon  ALP ALP
South-West Coast  LIB LIB
St Albans  ALP ALP
Sunbury  ALP ALP
Sydenham  ALP ALP
Tarneit  ALP ALP
Thomastown  ALP ALP
Warrandyte  LIB LIB
Wendouree  ALP ALP
Werribee  ALP ALP
Williamstown  ALP ALP
Yan Yean  ALP ALP

Tuesday 8 November 2022

US Midterm Medley - Senate Showdown

Finally, we turn to the US Senate which, as usual, is electing one-third of its members (in this case, the Class 3 senators) in a general election. However, what is not usual is that there are two special elections – Oklahoma’s Class 2 senator and California’s Class 3 – and because California’s Class 3 election would have occurred on November 8 regardless, this means there are two elections for California – one beginning immediately, and another for the term beginning in January. Unsurprisingly, the candidates in both California elections are the same, meaning (weirdly) there are 36 elections for 35 senators. Of course, these two races can give different results, but this would require some peculiar voting patterns across a meaningful number of voters, so we will assume that we only need to address 34 races, and we can treat California as one contest.

The remaining senators who are not up for election include 34 Democrats, 29 Republicans and 2 Independents who caucus with the Democrats. This means the Democrats need 14 seats to have control with the Independents and rely on the Vice President to break ties, 15 to avoid most of those ties, 16 to not rely on Independents and 17 to not need the VP either. So anything less than 13 Democrat senators in this election is a win for the Republicans, which means taking 22 races for themselves. They come into this with 21 senate seats on the line, though, so this is a net gain of one seat.

The Races

As with the Gubernatorial elections, several sources have already made predictions, and where these all agree, there is little sense in running the same numbers they have. This gives us forecasts for 27 races, leaving Republicans just three races from claiming a majority with eight options on the board.

STATE INCUMBENT MEDIA PREDICTION
Alabama Republican Republican
Alaska Republican Republican
Arizona Democrat Mixed
Arkansas Republican Republican
California Democrat Democrat
Colorado Democrat Mixed
Connecticut Democrat Democrat
Florida Republican Republican
Georgia Democrat Mixed
Hawaii Democrat Democrat
Idaho Republican Republican
Illinois Democrat Democrat
Indiana Republican Republican
Iowa Republican Republican
Kansas Republican Republican
Kentucky Republican Republican
Louisiana Republican Republican
Maryland Democrat Democrat
Missouri Republican Republican
Nevada Democrat Mixed
New Hampshire Democrat Mixed
New York Democrat Democrat
North Carolina Republican Republican
North Dakota Republican Republican
Ohio Republican Republican
Oklahoma Republican Republican
Oklahoma* Republican Republican
Oregon Democrat Democrat
Pennsylvania Republican Mixed
South Carolina Republican Republican
South Dakota Republican Republican
Utah Republican Republican
Vermont Democrat Democrat
Washington Democrat Mixed
Wisconsin Republican Mixed

[Source]

Colorado, Washington

All sources plant these two races as Democrat wins, except for the unclear political predictions of Real Clear Politics which leave these as tossups. Both have multi-term Democrat incumbents recontesting their seats who won the last election with percentages in the high 50s. In both states, despite a narrowing lead in recent times, the Democrats have never lost a poll with the single exception of a hypothetical ‘Incumbent v generic Republican’ question in Colorado ten months ago. Therefore, it seems safe to call these for the Dems.

Democrat

New Hampshire

Similarly, New Hampshire is broadly predicted for the Dems, with only Real Clear Politics and Politico dissenting to mark it as a tossup. Again, the polling has narrowed, but the Democrats have won the majority of these by far.

Democrat

Wisconsin

Wisconsin is the last senate seat where there is a majority view among media sources. Five sources mark this as Republican, the other four as a tossup. Democrats led the polling until August, even among right-wing sources like Trafalgar and Fox, but since then, blue leads have been few and far between against the two-term Republican incumbent. The hesitancy to write this off has likely been due to strong Democrat performances in Wisconsin, including a blue sweep in 2018, but this is not translating to the polls this year. (For comparison, Democrats led every poll ahead of the 2018 senate election).

Republican

Pennsylvania

Perhaps the most publicised contest of 2022, this race pits TV’s Dr Oz against Lieutenant Governor John Fetterman. Early polling was impressive for the Democrats but has slipped to give Oz a slight lead in the polls since October 26.

Sabato has this race leaning R, which aligns with the latest polling, but I’d also be wary about relying on numbers from the last fortnight after months of contrary results. What is apparent is that the number of undecideds has (naturally) narrowed. If this indicates that undecideds tend to fall out in Oz’s favour, things will only worsen for Fetterman. I’ll call this Republican, but this is the race I most expect to get wrong.

Republican

Arizona

The Cook Partisan Voting Index marks Arizona as naturally Republican, but the incumbent is a Democrat, and four of the nine media sources are colouring this one blue. The polling was remarkable for Democrats until the withdrawal of Libertarian candidate Marc Victor, and since then has still been strong, with only three of the subsequent 13 polls favouring the Republicans. There is some division in the data aggregates, with RealClearPolitics favouring the Republican while FiveThirtyEight and 270ToWin favour the Democrat, but the raw data is convincing enough for me.

Democrat

Nevada

Nevada polling is all over the shop, but with a preference towards the Republicans. However, Sabato was the only source willing to wager a bet on Nevada, and they had it as leaning Democrat. The Cook Partisan Voting Index has this as a Republican state, but the incumbent is a Democrat. In the closest race of 2022, a couple of factors are holding the Democrats back that will likely be decisive: Republican economic issues seem to be cutting through more than the Democrats’ focus on abortion rights, and Biden is so unpopular in the state that they sent Bill Clinton in to support instead.

Republican

Georgia

Sabato and FiveThirtyEight are both predicting the GOP to win in Georgia, the only calls across the senate races to go against an incumbent. Polling supports this call with the Republican candidate ahead in all aggregations and not losing a single poll in November.

Republican

Summary

That comes to 23 Republican and 12 Democrat senators elected in 2022 for a 2023 term of 52 Republicans, 46 Democrats and two independents. In short, Republicans are predicted to control the Senate and block much of Biden’s agenda for the second half of his presidency.

STATE INCUMBENT MEDIA PREDICTION
Alabama Republican Republican
Alaska Republican Republican
Arizona Democrat Democrat
Arkansas Republican Republican
California Democrat Democrat
Colorado Democrat Democrat
Connecticut Democrat Democrat
Florida Republican Republican
Georgia Democrat Republican
Hawaii Democrat Democrat
Idaho Republican Republican
Illinois Democrat Democrat
Indiana Republican Republican
Iowa Republican Republican
Kansas Republican Republican
Kentucky Republican Republican
Louisiana Republican Republican
Maryland Democrat Democrat
Missouri Republican Republican
Nevada Democrat Republican
New Hampshire Democrat Democrat
New York Democrat Democrat
North Carolina Republican Republican
North Dakota Republican Republican
Ohio Republican Republican
Oklahoma Republican Republican
Oklahoma* Republican Republican
Oregon Democrat Democrat
Pennsylvania Republican Republican
South Carolina Republican Republican
South Dakota Republican Republican
Utah Republican Republican
Vermont Democrat Democrat
Washington Democrat Democrat
Wisconsin Republican Republican

In 2024, 23 of the 33 senate seats up for election are currently held by Democrats or the two Independents who vote with them. That leaves little room to pick up the three seats needed to regain control. Therefore, a Republican win, as predicted here, is likely to also hold for the first half of the next Presidential term—good news for a Republican president but crippling for a Democrat.