the numbers guy

Author Archive

Canadian federal politics appear to be entering a period of significant change, with serious challenges facing all three major parties. That’s according to a series of recent polls, including a Forum Research poll today which suggests that a Justin Trudeau-led Liberal Party would win 41% of the vote and an overall majority in the House of Commons:

Forum Poll 8 Feb 2013 with Trudeau

What’s even more striking is the difference once Mr Trudeau’s name is removed from the poll wording – the very same Forum Research poll had figures of Lib 30% Con 32% NDP 26% when Mr Trudeau is not mentioned – a remarkable difference of 11 points and 78 seats for the Liberals:

Forum Poll 8 Feb 2013 without Trudeau

Of course all polls come with health warnings and this one may be somewhat understating NDP and Conservative support, while overstating Liberal support – at least according to the most recent ThreeHundredEight federal polling average.

However, the opinion polls are not the only evidence that, assuming Mr Trudeau wins the leadership, the Liberal Party could be looking at a significant boost in the polls heading into the summer. Mr Trudeau has also been drawing large crowds in places such as Kamloops in BC that are not exactly traditional Liberal turf.

600 people to see justintrudeau in #Kamloops, BC. Six. Hundred. People. For a #Liberal. #lpcldr

— Gerald Butts (@gmbutts) January 22, 2013

With the Liberal leadership election taking place on 14 April, it will be some time before we know whether the electoral promise of a Trudeau leadership matches up to the reality. In the meantime, it’s a sure bet that Messrs Mulcair and Harper will be working hard to shore up their own support, particularly on the Liberal-facing flanks of their support bases.

The next federal election is still two and a half years away, and there will be plenty of water under the bridge by then. What does seem clear is that those two and a half years will be anything but dull!


Season’s Greetings and happy holidays to all my readers.

Blogging here has been sparse recently – this is due to the fact that I’ve recently moved home, not just across the country but across continents. I am now living in Vancouver in the beautiful province of British Columbia, Canada.

Normal service shall resume in the new year – in the meantime, enjoy that eggnog!

It’s election day, and this is my best guess (not a projection) as to how tonight’s result will pan out. The state I am least sure about is Virginia (followed by Colorado):

In a nutshell, with just hours to go, Barack Obama is the clear favourite to win. The reason can be summed up in a single word – Ohio. Mr Obama has maintained a modest but consistent lead in polling in Ohio, a state whose 18 electoral votes are likely to prove decisive.

Without Ohio there are almost no viable routes to 270 for Mitt Romney – Mr Obama on the other hand can win without Ohio, although losing it would make his path to victory far more precarious. My ‘best guess’ map above is an example of this – Mr Obama would win that map even if he lost Ohio, but he couldn’t afford to lose Colorado or even New Hampshire.

Should make for a fascinating election night – if Ohio’s on a knife-edge, it could be a real nail-biter.

(PS – you can create your own ‘best guess’ electoral map at

Canadian federal politics got rather more exciting this week, as Papineau MP Justin Trudeau confirmed that he will enter the Liberal leadership race. Mr Trudeau will formally announce his entry at an event in his riding on Tuesday, and he is already the prohibitive favourite to win the leadership next April.

Forum Research have been in the field to ascertain the potential impact of a Trudeau-led Liberal Party on the Canadian federal scene. The remarkable results of their opinion poll will surely have alarm bells sounding among NDP and Conservative strategists – according to Forum, a Trudeau-led Liberal Party would win 39% of the vote, the Conservatives 32%, and the NDP just 20%.

Now this is just one poll, taken at a time when it’s simply too early to tell what impact Mr Trudeau will have on federal politics should he win the leadership next April. For example, Mr Trudeau is already facing pressure to set out his stall more clearly on major national issues – and it may be that, when he does so, his support levels decrease.

That said, while we should never read too much into a single yet-to-be-corroborated poll, a possible swing of this magnitude – with the Liberals being catapulted from a distant third to a strong first place – is really quite remarkable.

Instinctively, it’s difficult to avoid the conclusion that, with this much smoke there’s surely a fire out there somewhere. Perhaps the potential Liberal swing won’t be quite so large as this Forum poll suggests, but I would be surprised if future polls do not confirm at least some increase in Liberal support under a future Trudeau leadership.

I ran a seat projection based on the Forum poll to get a better idea of the potential impact a Trudeau-led Liberal party would have on the future shape of the House of Commons:

Forum’s headline party support figures of Liberals 39%, Conservatives 32% and NDP 20% would completely transform the shape of the House. An election contested on the existing 308 ridings would result in a likely Liberal minority government – with the Liberals winning 130 seats (+96 seats from 2011), the Conservatives 115 (-51), the NDP 56 (-47), the Bloc 5 (+1), and Others 2 (+1 – Elizabeth May and André Arthur).

Broken down by province, the Liberals would win big in Ontario, taking 64 seats (+53) to the Conservative’s 33 seats (-40) and the NDP’s 9 (-13). In Québec the NDP lead with 36 seats (-23) to the Liberal’s 25 (+18), the Conservative’s 8 (+3), the Bloc’s 5 (+1) and Others 1 (+1).

Elsewhere the Conservatives retain the lion’s share of the seats west of Ontario – taking 67 seats (-5 from 2011) to the Liberals’ 16 (+12) and the NDP’s 8 (-7). In Atlantic Canada, the Liberals would take 22 seats (+10) to the Conservative’s 7 (-7) and the NDP’s 3 (-3).

The overall projection assumed a uniform national swing, so the projected provincial results should be taken with a bit of a pinch of salt. Nevertheless, these results run counter to the commonly held view that a Trudeau-led Liberals would be a greater danger to the NDP in Québec than to the Conservatives in Ontario.

These numbers suggest that Mr Trudeau poses a very serious threat to the Conservatives in Ontario, which is shaping up as the key battleground – particularly given that Ontario will have 15 extra seats by the time the next federal election rolls round in 2015. Conservatives tempted to rub their hands in glee at the prospect of a Liberal -v- NDP slugfest in Québec should think again – their majority depends on holding their ground in Ontario where their position looks really quite precarious.

I look forward to seeing whether future opinion polls corroborate the results of this Forum survey – we could be in for a fascinating couple of years in Canadian federal politics.

Three days out, the result of Tuesday’s Québec election remains impossible to call. We’ve seen a flurry of polls over the past week, all of which indicate the Parti Québecois holding on to a lead of between 4 and 5 points. With the single exception of a 20 August Forum poll, the polling evidence is remarkably consistent about party support levels – the PQ on 32-33%, the PLQ on 26-28%, the CAQ on 27-28%, Québec Solidaire on 7-9%.

A seat projection based on these polling averages yields a fascinating result – the PQ are on a projected 67 seats, just over the 63 they need for a majority:

  • Across the province, the PQ win a majority with 67 seats, the PLQ win 33 seats, the CAQ 23 seats, and Québec Solidaire win 2 seats.
  • In Montréal, the PLQ are projected to win 20 seats, followed by the PQ with 18 seatsQuébec Solidaire with 2 seats and the CAQ with 1 seat.
  • In Québec City, the CAQ are projected to win 7 seats, with the PLQ and PQ winning 2 seats each.
  • In the Rest of Québec, the PQ are projected to win 47 seats, with the CAQ winning 15 seats and the PLQ 11 seats.

What makes the end result so difficult to call are the large number of closely contested seats. No fewer than 25 of the 125 ridings have a projected margin of victory of under 5%. Of these 25 close races, the PQ lead narrowly in 14, the PLQ lead narrowly in 8 and the CAQ in 3:

The PQ are either ahead or within 5 points of the leader in 24 of the 25 close ridings – meaning they could win as few as 53 seats or as many as 77. The PLQ are projected to win between 25 and 39 seats, the CAQ between 20 and 32 seats and Québec Solidaire 2 seats with an outside chance of a third.

There are still three days to go, and Alberta showed us that a lot can happen in three days. I’ll be keeping a close watch on the final polls of the campaign for any signs of a last minute Alberta-style swing to one of the parties.

With so many ridings this close, a late swing of just a percentage point or two could make all the difference. It’s all shaping up to be a very exciting election night on Tuesday.

Québec voters will go to the polls on 4 September to elect their new provincial government. The opinion polls had been quite volatile during the months leading up to the election, but have settled down somewhat since the start of the campaign – although they have pointed to a recent rise in support for the Coalition Avenir Québec. Recent polls also agree that the Parti Québecois, led by Pauline Marois, enjoys an edge of roughly five points over the second place Parti Libéral (led by Premier Jean Charest) and the third place CAQ.

Also contesting the election are the left-wing Québec Solidaire which polled 3.8% in 2008, and succeeded in winning its first ever seat in the Assemblée Nationale du Québec. They are a factor in a number of ridings on the island of Montreal, and will be hoping to add a second seat this time round. Also in the picture is the small, sovereigntist Option Nationale led by Jean-Martin Aussant who will be hoping to do well in Nicolet-Bécancour.

There have also been a number of boundary changes since 2008, which added additional seats to the regions surrounding the Island of Montreal – Laurentides-Lanaudière, Laval and Montérégie. The Chaudière-Appalaches, Bas-Saint-Laurent and Gaspésie–Îles-de-la-Madeleine regions lost seats to compensate.

According to my mid-campaign projection, here’s how the recent polling evidence would translate into seats:

  • Across the province, the PQ win a majority with 69 seats, the PLQ win 34 seats, the CAQ 20 seats, and Québec Solidaire win 2 seats.
  • In Montréal, the PLQ are projected to win 20 seats, followed by the PQ with 18 seats, Québec Solidaire with 2 seats and the CAQ with 1 seat.
  • In Québec City, the CAQ are projected to win 7 seats, with the PLQ and PQ winning 2 seats each.
  • In the Rest of Québec, the PQ are projected to win 49 seats, with the PLQ and CAQ winning 12 seats each.

The PQ clearly have their nose on front and, as things stand, are on track for an overall majority. However, with over a fortnight’s campaigning to come, there are far too  many close races that this election remains impossible to call.

The PQ will be looking to maintain their roughly 5-point edge over the PLQ and CAQ – if that lead narrows, the seat projection would change dramatically – and prospects for a PQ majority would shrink accordingly. This election is far from over yet.

There are now less than four months until the US presidential election in November. Presidential elections nowadays have such a lengthy lead-in (years not months) that it is almost startling to realise that election day itself is now just round the corner.

You just have to look at the final few months of the 2008 election to know that lots can change between now and election day. Nevertheless, it seems like a good time to check in and look at the state of the campaign.

At a media briefing last December, chief Obama strategists David Axelrod and Jim Messina outlined five viable paths to winning 270 electoral votes, assuming Mr Obama holds the states won by Mr Kerry in 2004 – states now worth 246 electoral votes. These paths are:

  • A ‘Western Path’ – Colorado (9 electoral votes), New Mexico (5), Nevada (6) and Iowa (6) = 272
  • A ‘Florida Path’ – Florida (29) = 275
  • A ‘South Path’ – North Carolina (15) and Virginia (13) = 274
  • A ‘Midwest Path’ – Ohio (18) and Iowa (6) = 270
  • An ‘expansion path’ – by winning Arizona (11), along with Colorado, New Mexico and Nevada (20 in total), Mr Obama could compensate for losses in Pennsylvania (20) and New Hampshire (4).

Seven months of campaigning later, how are each of these paths shaping up as viable routes to winning the electoral college?

Here are the current FiveThirtyEight polling averages for the four ‘Western Path’ states:

These numbers are not quite so promising for Mr Obama as they may seem at first sight. Recent polls in Iowa have the race as a dead heat – Mr Obama’s lead in the Iowa average reflects a two month old PPP poll that had him with a 10 point lead.

His lead in New Mexico is impressive although the state hasn’t been polled since late April. The Nevada figures are more recent, although the most recent poll in the Silver State is almost a month old. By contrast, we’re now getting polls out of Colorado most weeks, and the Obama campaign clearly enjoys a slight edge there.

Florida is another state now being polled most weeks, and here too Mr Obama has a slight advantage – though still too close for comfort:

Of the available routes to 270, it is the ‘South Path’ – winning both North Carolina and Virginia – that now looks most challenging for the Obama camp:

Just one of the five most recent North Carolina polls has Mr Obama ahead, and polls since mid-May have given Mr Romney consistent leads of anything from 1% to 8% .

The situation for Mr Obama in Virginia looks more promising and, indeed, he would lead by more in the average were it not for a recent We Ask America poll that had Mr Romney up by 48-43.3 – numbers which stand out quite strikingly from many months of polls favourable to Mr Obama in the state. This poll, along with the 6 May PPP poll in Iowa, looks like it may well be an outlier.

The ‘Midwest Path’ requires a win in the all-important state of Ohio and here Mr Obama appears to be doing pretty well:

It is difficult to understate the importance of Ohio in the electoral map arithmetic. Despite a string of disappointing monthly jobs figures in this high unemployment state, Mr Obama has consistently polled well here in recent months. That said, two of the four most recent polls had Mr Romney with his nose in front and four months is an awful long time in politics. We can expect a ferocious contest in this most crucial of battleground states between now and November.

Finally, there’s the ‘expansion path’, which would see the Obama campaign seek to win Arizona’s 11 electoral votes, along with those of neighbouring Colorado, New Mexico and Nevada. Mr Obama might need Arizona’s 11 electoral votes to compensate for losing a Kerry 2004 state such as Pennsylvania or New Hampshire.

Arizona now looks pretty much out of reach for the Obama campaign, who haven’t led in a poll there since mid-April. On the up side for Mr Obama, Pennsylvania and New Hampshire both look like terribly big asks for Mr Romney, even if they are not quite yet a done deal for the Obama camp.

Taken as a whole, these state-by-state polling figures show that the Obama campaign still enjoys a choice of viable paths to victory. Of the scenarios sketched out by Mr Axelrod and Mr Messina last December, only the expansion path (Arizona) now seems definitely out of reach, although recent polls out of North Carolina are making the southern path look increasingly challenging.

However, the paths are not, of course, mutually exclusive – an Obama win in Virginia for instance would more than compensate for, say, a stumble in Iowa along the ‘Western Path’.

At this stage, four months out, it’s difficult to do more than provide a snapshot of how things stand. Overall, it seems like Mr Obama retains a slight overall edge – a position he has enjoyed since the start of the year. However, with the conventions, debates and the rough-and-tumble final weeks of the campaign still to come, it’s simply impossible to predict what will happen in November.

  • None

Twitter – numbersguyblog