Thursday, November 27, 2014

Newfoundland prediction map

Using the figures I worked out in my last post, I've done up a map. I want to note I ended up using 3 as the Liberal figure as after looking at polls, it seemed more in line with reality.

While this would be terrible for the Tories, it would leave them with a few seats, and their leader, in the legislature.

The Trinity riding - the Purple TCTC (Too Close To Call) riding in the "rural" area - is so because of a high NDP vote that I expect to go partly to the Liberals. Given Liberal strength in Rural NL, it is quite likely this will go Liberal on election night.

The purple one outside the inset but near St. John's - Harbour Main - is so for similar reasons, but could remain blue.

The purple one inside St. John's - Cape St. Francis - is also so for similar reasons, but due to a very high NDP vote (38%) is also likely to go Liberal.

This would give the Tories 4 seats.

The NDP meanwhile should be able to hold on to their seats, with the possibility of losing the upper-leftmost seat - St. John's East - to the Liberals.

The remaining NDP seats are St. John's Centre, the "downmost" seat, and Signal Hill-Quidi Vidi, the Leader's seat.

The remaining PC seats are:
Topsail - the  "leftmost" riding on the St. John's map, where the Premier sits. The Tories won last time with 68% of the vote.
Mount Pearl North - The " rightmost" riding on the St. John's map, where the Tories won with  73% of the vote.
Ferryland - On the main map - where the Tories won with 72% of the vote.

Mount Pearl North was last held by the Liberals in 1993, when, in that year, it was won by a PC candidate.
Topsail was last held in 1996
Harbour Main in 1999
And Ferryland in 1971. In fact, the Liberals did not break 33% in the riding in the past 2 decades.

Note that all the Non-Liberal victories depend completely on the incumbent running again. If he does not, what happened in Humber East, where the Tories took 78% yet lost the by-election by 20 points, will happen again.

Wednesday, November 26, 2014

Quick post on rough estimates

This is partly in reply to a 308 post about the by-elections provincially in Newfoundland.

What I want to discuss is two things. First; what to do when you want to do a projection but are short on time. Right now I am very short on time (sorry for not posting more often!) so I'm not only truncating this post as much as possible, but I will teach you how to quickly do estimates.

First, note the PC and NDP numbers.

For the NDP you'll note 7.8 and 5.4, but more importantly, the "down by" numbers of 5.5 and 8.8.

Lets very crudely and roughly estimate and round that into 4 and 9 VS 7 and 5. That all averages out to somewhere in the ballpark of 7 and 7. My point?

A rough estimate shows us the NDP has lost half their vote.

Now look at the PC numbers. You'll note the "down by" is higher than the "vote won", but both numbers are close, within 4 points of one another.

Again, a rough estimate shows us the PC Party has lost half their vote as well.

Lastly the Liberals. You need to actually do a bit of math here, but I'll give you the spoiler; the Liberals are up by between 2.5 and 3 times their vote last election.

I personally think 2.5 is more realistic.

While I don't have time to go riding by riding and multiply all results by this factor, I read the post, did the math, and wrote this post in a grand total of 6 minutes.

Just because you are pressed for time does not mean you should avoid or ignore making a projection!

Wednesday, November 19, 2014

Are By-Elections a Sign of Things to Come in a General Election?

A couple days ago, the Conservatives held on to two of their seats in Yellowhead and Whitby-Oshawa by fairly strong margins (42.6% in Yellowhead and 8.5% in Whitby-Oshawa). At least to me, this was not unexpected; Yellowhead did not feature the hullabaloo that Brandon-Souris or Fort McMurray did, so the Liberals were going to do well (they did, winning just under 20%, their best since 1993) but they were never going to come close to winning. In Whitby-Oshawa, the Liberals came closer that I expected, but I didn't think they would win - the combonation of this being Flaherty's old riding, his wife Christine Elliott's influence, and the fact that the Conservative candidate, in addition to being a climate change denier, is also a former two-term mayor of Whitby... yeah, that was just too much to overcome. Still, like in Yellowhead, the Liberal candidate managed to increase the party's share of the vote by quite a bit, going from 14.1% in 2011 to 40.7% Monday night.

The somewhat understated story of these by-elections (well, in priority behind Trudeau anyways) is the impact of the New Democrats in the various ridings... which is to say, there has been no impact. They have not come close to picking up any of these ridings from other parties, while losing one of their own seats and coming perilously close in another (the third was a mixed bag). As the chart below shows, the NDP have lost some level of influence in almost every single by-election in 2011.

Mulcair has been questioned on these results, and his response was pretty pat: "... by-elections are not always a great indicator of the general (election)," and that "... we've got a lot of work to do but we also know that our numbers have never been better heading into a federal (election)."

Well, I will give Mulcair the last one: their polling numbers have never been better when going into a general election, when you put those numbers up against previous election results. The problem, of course, is that the NDP wasn't the official opposition in any of those previous elections, and the current results are far below their 2011 numbers in every region of the country, including (if not especially) Quebec. So, yeah, I guess technically he's right, and there is the caveat of "the campaign matters," but as we've seen so far, any actual campaigns the NDP has been in have not exactly gone well.

His other main point is that yes, by-elections by themselves are not indicators of results in general elections. But it is also pretty hard to argue against trends, and there is an obvious trend that Mulcair is purposely ignoring. For example, here is the same chart above for the by-elections between 2008 and 2011:

With the exception of Winnipeg North, there was no total loss for the NDP - they either maintained their position or ended up higher than in 2008, even if they lost votes and percentage., in all other ridings, which the Liberals (at the time the Official Opposition) failed to do during the same period.

So yes, by-elections are a singular event, and they're more prone to local influence than any normal general elections... but come on. A party that is the Official Opposition, that says momentum is on their side and that the third party are has-beens, should be able to put up better results than this. At best, the NDP should be able to keep the Liberals down in obvious uphill battles like rural Conservative ridings or their own downtown core (or, for that matter, the Greens in Victoria), but that has simply not been the case.

And then there is the whole matter of feeding into the various narratives. What the public sees between the two main opposition parties is a battle over who can defeat Harper, and these by-elections clearly show that the NDP are unable to do that. Do you think that isn't going to affect how people vote outside of the ridings holding by-elections? Momentum, such as it is, is on the side of the Liberals, and at best the NDP can hope to hold to their 20-25% of the vote if the trend holds. That is not what I call a "never been better" proposition.

Wednesday, November 5, 2014

US Election quickie

Teddy here; as most of you likely know by now, US politics is not my cup of tea. However, I thought I should at least outline some results from last night's Mid-Terms.

R - 53
D - 47
This presumes that Angus King will caucus with the majority. If he does not, we are looking at a balance of 52-48. Regardless, Republicans now control the Senate.

A slight gain for the Republicans.

Another slight gain for Republicans.

In the more local races

T-1 (exact tie)


There are still 5 chambers where the result is unclear.

In terms of Trifectas; that is, where one party holds the House, Senate, and Governor's position

As usual in US elections, a number of Ballot measures, IE Referendums, were held.

California has adopted a "rainy day fund" similar to that used by Saskatchewan. It's a brilliant idea, frankly, that forces governments to save during the good times to fight deficits during the bad times.

Minimum wages in Arkansas, Nebraska, Illinois, South Dakota, and Alaska were voted an increase.

Oregon, DC, and Alaska have all legalized Marijuana, while Florida refused to legalize even medical marijuana.

Lastly, Tennessee, famous for having no income taxes, voted to ban the state constitutionally from implementing one.

Good websites for more information include Ballotpedia and the NY Times election site.

Tuesday, November 4, 2014

On creating good maps

Teddy here. I am still getting used to my new computer, so the graphics quality of my maps may be below par for a while, however, I take pride in creating maps that can share quality information.

Take for example the Summerside election. Summerside is the second largest city in PEI and where I grew up. Basil Stewart has been mayor there since 1985 and has seen the city though it's worst times and best times. Stewart is a known Tory, having run to be a PC MP in 1993. He said publicly this would be his last term and last election. Normally, Stewart can be expected to win by huge margins.

Last night, he lost.

And not just by a bit, by a huge margin, 69% to 31%.

Any map you'd thus see of Summerside would be expected to be a single colour, as, Bill Martin, the Mayor-Elect, won every ward. However, I do not consider this to be a good map. It does not really share any interesting information.

So what is a good map? Frankly, this:

Shown in Blue are wards that Stewart managed to get over 31% in, 31% being the city-wide average of vote he obtained. Martin won all of these wards by a huge margin, but any map that shows the "winner" of each ward would miss this divide in the city.

You can see in a darker outline, the "downtown" Summerside; which was by in large the original city prior to municipal merger in the 90's, IE the city that originally elected Stewart. If you want to compare Summerside to Toronto (and this is not the best idea, but one can stretch it for an analogy) ward 4 and 6 are like the "old Toronto" and the remainder of the wards like the "Inner Suburbs"

Martin, as far as I understand, is from the Wilmot area, this being, the areas in wards 7 and 8. With that in mind, this map shows the "urban" and "suburban" split, if you can call it that.

This map reveals a vote pattern that could be important in Provincial and Federal elections, where Tories and Liberals might want to target resources.

I uurge everyone who makes maps to keep in mind that you are not stuck to the way things usually are or the way things always are. Think outside the box. This will allow you to expand your set of tools and make better maps.

Wednesday, October 29, 2014

Alberta By-Elections

Teddy here on his laptop with a quick update.

The above is a colour coded results table.

The largest victory is in Calgary-Foothills.
The ratio margin is 0.5129
That means even if the PC Party had only managed 51.29% of their vote they would have still won (or at least tied) in the riding.

Prentice ran here. This indicates that he has some personal support over and beyond his party, which only makes sense. This was the least-likely loss for the party and would have been most damaging if they had lost it.

Number Two is in Edmonton-Whitemud.
While the ratio is 0.5236 there is a caveat here.

Not only is this the best NDP result, but it's also the best Liberal result, and, the worst Wildrose result. This is an Edmonton riding. Edmonton is more to the left, and these results indicate to me that the Tories can still win on the splits in the city.

Up next is Calgary-Elbow.
The ratio margin here is 0.8102, a much closer race.

There were two spoilers here, the first is that this is the riding of Redford, and the second was the very strong Alberta Party candidate. My personal read on this is that the AP vote is the "I want to vote PC but what they are doing recently upsets me" vote. This vote is available across the province for the party best able to pick it up. Note that if Wildrose or the Liberals had picked up this vote, they would have won the riding. This is good news for both parties; if they can tap in to the anger at the PC Party, and hold on to it (without losing it to fear as Wildrose did last time) they could win the election. The problem here is that neither party has been able to do so, and these results simply show that the PC Party is still flying high.

Last on our list is Calgary-West.
The ratio margin is a small 0.9347.
This means even a loss of 7% of the vote would have resulted in a loss for the Tories.

The simple fact that this is the last riding on the list is important, it's the only one where the Tories were not running a Cabinet minister. In effect, it is the riding the Tories decided they could afford to lose. They did not lose it.

For the reasons outlined above, this is great news for the Tories, bad news for the Liberals, NDP, and Alberta Party, and Terrible news for Wildrose. The latter needed Calgary West. Losing it will have repercussions.

Monday, October 27, 2014

Remember to vote, and updates

I hope everyone who follows this blog who can vote today has already done so. Those who have yet to, I hope you do so soon, even if you simply submit an empty ballot.

I've voted for Morgan Baskin. I am unconvinced she could do any worse than the "big 3" candidates. I also want to encourage her to stick with politics as we need more people like her.

As for updates.

In Tunisia, the Liberals (Nidaa) have secured a victory. The Islamists have conceded defeat.

In Ukraine, now that more votes are in, I've updated the results. With 75% of the vote in, things stand as follows:

As noted in the graphic, the Front and the President's party are setting up for a coalition government.