Thursday, March 06, 2014

Building a seat model for New Brunswick

The idea behind a seat model is to try to convert the popular vote estimates from an opinion poll into a seat count as the latter is far more relevant to our politics.

The fact that Richard Hatfield lost the popular vote in 1974, did not prevent him from being premier of a majority government for four years before winning the popular vote in 1978 and 1982. The fact that 40% of New Brunswickers voted for non-Liberal candidates in 1987, did not get any of them elected. More recently, Shawn Graham also won a majority government while losing the popular vote in 2006.

To summarize: the popular vote doesn't tell the whole story in our system of government. It is the seats that count.

The Method

There are two general approaches that one can take in a seat model. Uniform swing or proportional swing.

Uniform swing has been used for years in Britain and Australia.

It means that all parties take the same arithemetic benefit or loss from their change of support throughout the jurisdiction. So if the Labour vote is down by 10 percentage points from 40 to 30, it drops by 10 percentage points in every riding. In riding X, their support drops from 50% to 40% (which means they've lost 20% of their base votes), in riding Y it drops from 15% to 5% (which means they've lost 67% of their base votes).

The alternative is proportional swing. This is what is generally used by election projectionists in North America, including the site ThreeHundredEight.com which has projected a lot of Canadian elections from polls. Calgary Grit and FiveThirtyEight use seat models which are built on a base of proportional swing, with many other fancy features.

It means that parties move in support relative to their base. So if the Labour vote is down by 10 percentage points from 40 to 30, it drops by 25% in every riding, meaning they take a relatively bigger hit in ridings were they had more votes to lose. In riding X, their support drops from 50% to 37.5% (25% of 50% is 12.5 percentage points), in riding Y it drops from 15% to 11.25% (25% of 15% is 3.75 percentage points).

The teaser seat modelling I have done on Twitter has been based on a proportional swing method.

I have spent the past several weeks improving and testing the model in anticipation for the main event in September.

One of the things I tested was each of the past 5 elections' popular vote versus the actual results in seats.

Interestingly, I found that proprtional swing does a better job producing the overall provincewide seat total. However, it does this while making many errors at the local seat level which tend to cancel each other out. While the uniform swing doesn't do quite as will with the final overall result, it does do a better job at the local seat level (its errors tend to be in one direction rather than both, so its overall numbers are worse).

I played around with a number of "half-way" alternatives. The most simple of these, taking the result under uniform swing and under proportional swing and simply averaging them proved to work the best. On average it tied the uniform swing for accuracy at the local level, and proportional swing for accuracy at the provincial level. I will be using this hybrid model going forward.

This model uses results of the 1995-2010 general elections as a base. I have transposed the 1995-2003 ridings and the 2006-2010 ridings onto the new boundaries to be used for the 2014 election. The 1995-2003 are weighted half as much as the two more recent 2006 and 2010 elections to create an "average" election result from which the model builds its projection based on change in the popular vote from that hypothetical election to what the opinion polls are saying. For the Greens and PANB, their 2010 results are used as a base as that is the only election they have contested. Votes for defunct parties and independents are ignored by the model.

This model is not as fancy as the models used by Calgary Grit and FiveThirtyEight, which model thousands of iterations based on the margins of error, etc and generate a percentage liklihood of each parties' chance of winning a particular riding. I do run about a dozen scenarios through the model pushing each party to the upper and lower range of the margin of error to generate seat ranges.

Adjustments: A Hand Up for the Leader

Another measure that I have incorporated into the model is a "leader bounce".

I've looked at the result of the leader in his or her riding in those elections he or she contested, and that of his or her party in that riding in the elections immediately before and after his or her leadership tenure. This allows one to assume how a generic candidate of the leader's party might have done in that riding so that it can be compared to the leader's result.

For instance, Robert Higgins led the Liberal Party in the 1974 election. He ran in the riding of Saint John Park. He received 14.6 percentage points more of the popular vote in his riding than the Liberal Party did provincewide. We cannot look at the previous election as Saint John Park was part of the 4-member Saint John Centre riding at that time. We therefore look at the two succeeding elections. In 1978, the Liberals beat their provincewide showing in Saint John Park as well, but only by 2.6 percentage points. In 1982, the Liberals did 4.6 percentage points better in Saint John Park than they did provincewide. The average of these two comparable elections is a 3.6 percentage point lean to the Liberals, while the riding leaned 14.6 percentage points to the Liberals when Higgins was leader. We therefore calculate Higgins' leader bounce as 11.0 percentage points.

The analysis finds that 15 of the 17 men and women who have led one of the three major political parties into an election since 1974, have had a bounce on average. If we single out each election contest, of the 10 elections since 1974*, the 3 major party leaders have had a bounce 25 of 30 times. The two exceptions are Richard Hatfield, who underperformed what we assume a generic Tory would have done in Carleton Centre in 1978, 1982 and 1987 and Shawn Graham, who underperformed what we assume a generic Liberal (actually, not that generic as the comparator is his father) would have done in 2003 and 2006. Hatfield exceeded expectations in 1974, while Graham exceeded expectations in 2010. All other leaders have exceeded expectations versus their provincewide popular vote every time they've led their party to the polls.

So, it is clear to say that there is a leader bounce. Unpacking the numbers a bit, a few other potential conclusions can be observed:

  1. Liberal leaders tend to get smaller bounces than Conservative and NDP leaders. My conjecture on this is that because all 6 Liberal leaders since 1974 have run in ridings that lean Liberal to begin with, their potential voter pool is already somewhat saturated. Conservative leaders Dennis Cochrane and Bernard Lord ran in ridings that were then somewhat less Conservative-leaning than the province as a whole, as did NDP leaders John LaBossiere and Roger Duguay.
  2. Liberal and Conservative leaders tend to get their biggest bounce in their first election as leader, while NDP leaders tend to see their bounce grow over time.
  3. While Liberal leaders tend to have the same-sized bounce whether they win or lose the election, Conservative leaders actually do better in their home riding relative to the provincewide performance of their party when losing the election than when they win. Thus, it is more likely for a Liberal leader to lose his or her seat when being swept out of office than a Conservative, though this has not happened in recent history.

These conclusions should all be taken with a grain of salt. The whole analysis is based on 30 cases, which is a very small sample size. The three observations above are drawn from subsets of that already small sample. However, I believe that there is a compelling enough case to adjust for the presence of the leader. I have applied an "anti-leader bounce" to past ridings in elections where a leader was a candidate, and have in turn applied an appropriate leader bounce to the five ridings being contested to the party leaders in 2014.

*I thought I would explain why I only went back to 1974 for those who were interested, but it was too long a tangent to include in the main body of the post. Prior to that, analysis is difficult if not impossible as multi-member ridings were in place and calculating voting trends is very difficult. New Brunswick used the block voting system prior to 1974 in most ridings. Under this system, many ridings returned more than one member and each voter had as many voters as his riding had members. For instance, a voter in Gloucester County could cast 5 votes, while a voter in Campbellton could cast only 1. Thus, it is impossible with the data available to calculate the popular vote (that could only be done by examining every individual ballot). Moreover, as voters were not required to cast all of their votes (our Gloucester voter could have chosen to vote for 1, 2, 3, 4 or 5 candidates) and they could split their ballot (that is to say our Gloucester voter might have case 2 ballots for Liberals, 2 ballots for Conservatives and 1 for an NDPer). Due to this additional complication, it is impossible attempt to accurately transform the votes into a form that could compare with the popular votes in single member ridings after 1974.

Adjustments: A Hand Up for Incumbents

I've also done some research on whether or not the "incumbency factor" exists in New Brunswick elections. I looked at the 1978, 1982, 1987, 1991, 1999, 2003 amd 2010 elections (all elections where the previous election was fought under the same boundaries). In those elections, it showed that on average incumbents did about 6 points better than non-incumbents controlling for the regular partisan advantage within their ridings. This worked out to be virtually the same for Liberals and Conservatives. I also looked at the 1995 and 2006 elections to determine whether or not this holds true when the incumbent is running in a riding with significant portions that he or she has not previously represented. In these cases it looks like incumbents do on average 4 points better. I have applied a five point "bonus" to incumbents, including in cases where there are two incumbents in the same riding (meaning they have no relative advantage over each other, but they do over minor parties).

Effectiveness: How Would Have it Worked?

I plugged the popular vote results for the last 5 elections into the model as a test. It called the overall seat total correctly in the 2006 election, missed by no more than 1-2 seats per party in 1999 and 2003, and no more than 3 seats per party in 1995 and 2010. The incubmency factor was not in play in these tests and may have improved the results. In any event, if the polls are correct (always a big if), it seems this model should come fairly close to the seat totals as the election approaches.

I will begin using this seat model with the CRA poll to be released later today.

Monday, February 24, 2014

Whither the People's Alliance?

This is my latest assessment of where the five parties stand a good chance of winning in 2014. First, I assessed the general leanings of each of the new 49 ridings for the two main parties, then I looked at the NDP's best shots, and then the Greens. Now I turn to the People's Alliance.

Overview

In some ways, the People's Alliance provides the most difficult task for objective analysis. The NDP has been on the ballot for decades, so it is easy to look for trends and consistent pockets of strength. For the Greens, though they too have only fought one election, there are at least federal election results where the Greens have had consistent campaigns and trends going back to 2004.

The People's Alliance has no federal counterpart and has no provincial electoral history. And even in 2010, they only contested 14 of 55 ridings. How might they have done in the other 41 ridings had they been on the ballot? It is nearly impossible to say.

However, there is one other dataset at which we can look. That is the results for the Confederation of Regions party in the 1991 and 1995 elections. The People's Alliance does not seem to welcome comparisons to CoR. However, they are both populist parties made up principally of right-of-centre individuals with a skepticism towards at least some aspects of official bilingualism. Even if PANB is not a successor to CoR, they would seem to compete for the same universe of voters.

It is not likely a coincidence that the leader and deputy leader of the party both hail from areas where CoR won seats in 1991.

It is easy to write off the People's Alliance as a fringe party after they got only 1.2% of the vote in 2010. However, that does not tell the whole story. The People's Alliance only contested a small subset of ridings, PANB got 4.9% of the vote in the 14 ridings where they actually ran candidates.

Compare this to the Greens who took 5.0% in the 49 ridings they contested. When one looks at the potential for these two parties, they must be treated at least as equals. Indeed, the Alliance vote was more "efficient" in that their best riding saw their candidate take almost 50% more of the vote share than in the Greens' best riding.

Curiously, the People's Alliance ran in only 2 of the 9 ridings in Central New Brunswick which had been CoR's strongest region in 1991. This suggests there was untapped potential for the party in 2010. Indeed, with the PCs almost certain to lose some of their vote from 2010 (when they scored their second highest popular vote in modern history), the PANB is well positioned to capture right-of-centre voters who would not consider the Liberals, NDP or Greens.

Taking a look back to CoR's 1991 results, they won 8 ridings. I'll lay them out here and explain how they relate to current ridings:

  • Fredericton North: The current Fredericton North falls completely within the boundaries of this old riding, but significant portions of the old Fredericton North can also be found in Fredericton-York and Fredericton-Grand Lake
  • York North: This old riding is distributed almost evenly between Fredericton-York and York
  • York South: Most of this old riding can be found in York and Fredericton-Hanwell, small portions of it can be found in New Maryland-Sunbury and Charlotte-Campobello
  • Sunbury: This riding is split roughly in three between Fredericton-Grand Lake, Oromocto-Lincoln and New Maryland-Sunbury
  • Oromocto: This riding is wholly contained in the new riding of Oromocto-Lincoln
  • Southwest Miramichi: This riding is almost wholly contained in the new riding of Southwest Miramichi-Bay du Vin, small pieces of it are in Miramichi Bay-Neguac
  • Riverview: Most of this riding is still found in the modern riding of Riverview, with pieces of it in Albert
  • Albert: This riding is almost wholly contained within the new, larger Albert riding; a small piece of it can be found in Sussex-Fundy-St. Martins

One of the advantages that CoR had in 1991 which they won't in 2014, is that there were no PC incumbents. Right-of-centre voters had no sitting MLAs where they were naturally aligned. This probably allowed them to get higher vote totals than they might have otherwise. For this reason, PANB may want to target opportunities where there is no PC incumbent, such as New Maryland-Sunbury.

First Tier Targets

These are ridings where PANB has nominated candidates early, the candidate has some profile, and the riding has shown a sympathy for the party. All of these are key factors in a potential breakthrough.

Fredericton-Grand Lake

Party leader Kris Austin will run here. He scored an impressive 20% of the vote in the old Grand Lake-Gagetown riding in 2010. Even though there was no People's Alliance candidate on the ballot in Fredericton-Fort Nashwaak in that election, the redistributed boundaries give Austin a starting base of 15%. This is because Austin took a whooping 30% of the vote within the Grand Lake-Gagetown polls that move into this new district.

Charlotte-Campobello

This was the Alliance's second best seat in 2010, where St. Andrews mayor John Craig took nearly 7% of the vote. They have again nominated a candidate here with a high profile and by nominating her early that should give them an organizational edge. Working against them is the fact that this was not a strong area for CoR in 1991, when they placed third in both Charlotte West and St. Stephen-Milltown. However, this riding brings in polls from the McAdam area which gave CoR 54% of the vote in 1991. This is also a riding where the NDP has the potential to do well, so if the Alliance can make a good run here, it could become a four-way race which could be won with less than 30% of the vote.

Southwest Miramichi-Bay du Vin

CoR won the Southwest Miramichi part of this riding by a slim margin in 1991, while doing very poorly in the Bay du Vin portion of the riding. In 2010, Wes Gullison got 5% of the vote here - roughly mirroring the average PANB received in ridings where they stood candidates. Therefore, on paper this does not really look like a great riding for the People's Alliance. However, Gullison is already renominated and should have the advantage of an early start. This combined with his profile as deputy leader of the party, should give him a slight edge.

Second Tier Targets

These are ridings where either COR did well in 1991 or the People's Alliance did well in 2010 and where there is either no incumbent, or where the incumbent running has not represented the majority of the new riding previously.

New Maryland-Sunbury

This would have been a very strong riding for CoR in 1991 and has the advantage of being one of the few ridings where the People's Alliance can realistically compete where there is no PC incumbent on the ballot. They would be wise to nominate a candidate here soon.

York

York North was the Alliance's third best riding in 2010, and it is from that riding that the new York draws most of its population. Carl Urquhart, PC incumbent from the old York riding, has prepsented less than 20% of this new territory. The PANB's 2010 candidate was former Nackawic mayor Steven Hawkes. He did very well despite being nominated late. Nackawic finds itself in the centre of this new riding, while it was on the outer-edge of York North with some of its closest neighbours, including Hawkes' hometown of Canterbury in another district. Were Hawkes to run again and be nominated early, the stars are fairly well aligned for him, relatively speaking.

Gagetown-Petitcodiac

CoR would have won or nearly won this riding in 1991. The PC incumbent from Grand Lake-Gagetown has represented a minority of the riding.

Third Tier Targets

These are ridings where CoR did well in 1991 or the People's Alliance did well in 2010, but are represented by strong PC incumbents.

Fredericton-York

MLA Kirk MacDonald has served since 1999 and has won all elections easily, except for 2003 when he won by only 101 votes. The new Fredericton-York riding draws just under 50% of its population from MacDonald's current York North riding, however when one takes territory he currently represents as well as territory from the old pre-2006 riding of Mactaquac, MacDonald has represented a solid majority of this area. While this riding would have gone for CoR in 1991, it would have done so against first-term incumbent Liberal backbenchers, far easier targets for a conservative-populist than a four-term Conservative.

Albert

Unlike Fredericton-York, this riding is in fact largely unchanged from when CoR won it in 1991 (it has added a bit of Riverview and the village of Salisbury). However, as with York North and Fredericton North (the 'ancestors' of Fredericton-York), then the CoR candidate defeated a one-term Liberal backbencher. Defeating a four-term Conservative like Wayne Steeves is a far taller order.

Fourth Tier Targets

These are ridings where CoR did well in 1991 but the People's Alliance did not contest in 2010. They are also ridings where demographic changes mean CoR would not likely have done as well today as they did 20 years ago and thus the Alliance is less likely to find them attractive targets.

  • Oromocto-Lincoln
  • Fredericton West-Hanwell
  • Fredericton North
  • Riverview
Conclusion

The People's Alliance are likely longshots to win any seats if current polling trends hold. However, if they can manage to capture a significant slice of declining PC support and concentrate that in the Fredericton area as CoR did in 1991, it is not unrealistic for them to see a win. Kris Austin starts with a base of 30% in his home turf, if he can take 30% across the whole of his new riding then he would be very well positioned to win. It is not hard to imagine 35% of the vote being enough to carry Fredericton-Grand Lake in this election. I would say Austin's odds at winning are similar to the odds of Greens leader David Coon.

Overall, I would say that the People's Alliance are more likely to wind up as the fourth party in the legislature than the Greens. That, however, remains an unlikely scenario.

Monday, February 10, 2014

Whither the Greens?

This is the third post in my series on the upcoming New Brunswick general election. The first looked at New Brunswick's newly redrawn ridings. The second looked at where the NDP might breakthrough.

This post looks at places for a possible Green breakthrough, something that is far less likely than a breakthrough for the NDP. Next will come a look at potential breakthroughs for the People's Alliance.

Unlike the PCs, Liberals and NDP, the Greens (and PANB) have little electoral history to look at to judge where they might do well, as they have only contested one previous election. With only one dataset, it is impossible to know which areas of strength were due to a natural affection for the Green Party and its positions, and which were due to unique circumstances, like the strength of the local Green candidate, the weakness of other parties' candidates, a fleeting local issue, etc.

That being the case, this analysis should be taken with several grains of thoughts (even more than my usual observations).

Overview

The Greens seem to be stuck at sub-5% in the opinion polls. This would suggest that they are unlikely to win a seat under our first-past-the-post (FPTP) electoral system. However, it is not impossible. If their efforts are very specifically focused, they could succeed. One needs look no further than the example of federal Green leader Elizabeth May who won a seat in the last federal election with a nationwide popular vote of 3.9%. Other examples of success with this level of support would be the BC Progressive Democrats (1 seat on 5.7% in 1996), the Alberta Representative Party (2 seats on 5.1% in 1986), the Manitoba Liberals (1 seat on 7.5% in 2011), Quebec Solidaire (1 seat on 3.8% in 2008), the Quebec Action Democratique (1 seat on 6.5% in 1994), the PEI NDP (1 seat on 7.8% in 1996), and the Newfoundland NDP (1 seat on 4.5% in 1996).

When one looks at the data, the first thing that stands out is that the Green strength is found Moncton and southeastern New Brunswick. In 2010, 6 of the Greens' 8 best ridings were in the Moncton area; as were 10 of their 14 best. This aligns with the other data source we can consider: federal elections. Moncton—Riverview—Dieppe and Beauséjour have generally been the Greens best federal ridings since their breakthrough election of 2004.

In the 2010 election, under the new boundaries the Greens would have had 7 ridings where their vote more than doubled their provincewide popular vote. It is in these ridings that I'll focus my analysis.

Fredericton South

This would have been the Greens' strongest riding, scoring nearly 14% of the vote - well over 3 times their provincewide popular vote. It is therefore no surprise that Green leader David Coon is running in this riding. Adding to Coon's chances here is not only the strong Green base, but also the fact that this is the only riding in the province likely to see a 4-way race, which means someone could conceivably win the seat with less than 30% of the vote.

Memramcook-Tantramar

This would have been the Greens' next best riding. In fact, under the old boundaries Tantramar was the best riding for the Greens in 2010. Boundary changes make this area a bit weaker and Fredericton South a bit stronger. Like Fredericton South, there is a chance of a four-way race here, though it is somewhat less likely. The Liberals and NDP have not shown major strength here since there was a three-way PC/Liberal/NDP race in a 1998 by-election with each taking more han 30% of the vote.

Moncton South, Fredericton North and Moncton Centre

The Green Party would have taken more than 10% of the vote in each of these ridings in 2010. But without a four-way race, or a huge surge in Green support it is difficult to imagine the Greens taking any of these ridings. Fredericton North may be an outlier in this group as much of the Green strength here can be attributed to then party leader Jack MacDougall who is not expected to reoffer.

Moncton East and Moncton Northwest

The Green Party would have taken more than 9% of the vote in both of these ridings in 2010, which is better than double their provincewide share. Again, there is not much chance of four-way races here which is really what the Greens need to breakthrough at their current levels of support. In Moncton Northwest, a Tory stronghold, there is an outside chance of a moral victory for the Greens as they could place a distant second here under the right circumstances. That is highly unlikely however if the Liberals maintain the strong standing they've seen in recent polls.

Conclusion

At 5% of the popular vote, it would be a reach for the Greens to win any seats. However, there are 2 unique opportunities for the Greens. In Fredericton South and Memramcook-Tantramar the Greens have ridings which are both areas where they have a support base on which to build and potential four-way races that could allow them to sneak up the middle. The Greens would be well served to focus all of their resources on these two seats.

Wednesday, February 05, 2014

More fun with maps

I really can't shake the maps bug I caught a few days ago.

So here is another one I've been working on. Suppose Canada and the United States merged and each of the 10 provinces became states. The result would be a 60-state United States. There would now be 120 U.S. Senators (2 for each state) and the 435 members of the U.S. House of Representatives would have to be reapportioned among the 60 states. In presidential elections, each state would have the same number of electors as they have members of both houses of congress (so a minimum of 3, 2 senators + 1 congressman).

The territories, as is U.S. practice, would lose their representation in the Senate, get a non-voting delegate to the House and have no say in presidential elections.

Here is what the new U.S. electoral map would look like:

New Brunswick, Nova Scotia, Prince Edward Island, Newfoundland and Labrador, and Saskatchewan would all get the minimum 3 electoral votes in presidential elections, meaning that they would have one at-large congressman representing the entire province state, but 2 senators, the same as big states like California, Texas, Florida, New York, Ontario, Illinois and Ohio.

American states have to sacrifice 44 seats in the House of Representatives to make room for the 10 new ex-Canadian states. And their relative share of representation in the Senate drops from two percent to one-and-two-thirds percent. Specifically, the following states would lose representation under the current U.S. congressional apportionment formula to make way for Canadian seats in the House of Representatives:

  • California -6
  • Texas -4
  • Florida -3
  • New York -3
  • Georgia -2
  • Illinois -2
  • Michigan -2
  • Pennsylvania -2
  • Alabama -1
  • Arizona -1
  • Colorado -1
  • Indiana -1
  • Iowa -1
  • Kentucky -1
  • Maryland -1
  • Massachusetts -1
  • Minnesota -1
  • Nebraska -1
  • Nevada -1
  • New Jersey -1
  • North Carolina -1
  • Rhode Island -1
  • South Carolina -1
  • Tennessee -1
  • Virginia -1
  • Washington -1
  • West Virginia -1
  • Wisconsin -1

To win a presidential election 280 electoral votes would be required (up from 270), due to the electoral college growing by 20 due to new senators.

It would be increasingly difficult for Republicans to win the presidency. Polls suggest Democrats would be the heavy favourites in all provinces in presidential elections, even winning Alberta by about 40 percentage points. If one is generous, we could say the provinces would become 9 blue states and 1 swing state.

This would mean the Democrat would start with 259 electoral votes in the bag, to 173 for the Republican, with 126 swing votes. If one gives Democrats credit for Pennsylvania, a state they've not lost since 1988, they start with 277 votes and the Republican would have to win every single swing state in order to carry the election.

I sure do love fun with maps.

Monday, February 03, 2014

Fun with maps

On Saturday afternoon, the CBC's Dan McHardie tweeted out a cool map on Twitter. I went into fun-with-maps overdrive...

I haven't quite shaken the bug. Here are a couple more which I find highly interesting.

First Canadian provinces as American states with comparable GDP (PEI and the territories have economies far smaller than any U.S. state so they're not labeled):

Then Canadian provinces as American states with comparable GDP-per-capita ratios. This one was fascinating. Six of the 13 provinces and territories are at about the same level or richer than Delaware, the richest American state (as measured by GDP-per-capita). Even Canada's "poor" provinces New Brunswick, Nova Scotia and PEI hold up pretty well. New Brunswick lines up with California. Nova Scotians and P.E. Islanders are on average richer than Texans.

Fun with maps is almost as fun as fun with numbers!

Here is a revised version of the first map, with an explanation of what it is in case it finds itself floating in the internet without its explanatory tweet:

Monday, January 27, 2014

Whither the NDP?

The New Brunswick NDP has been on a roll of late.

They tend to get equal billing to the PCs and Liberals in the press - a status not enjoyed by their fellow non-seat parties the Greens and People's Alliance. And they're soaring in the polls; they eked out a statstically insiginifcant lead over the Tories in a recent poll, and have been consistently at or above 19% in every quarterly CRA poll since May 2011. The NDP has often had interelection spikes of support in New Brunswick, but none this big or this durable. It will be very interesting to see whether the NDP can hold on to all or most of this support in September, and what that might mean for them in terms of seats.

What would 20% of the vote look like for the NDP?

Well it is hard to say. Our first-past-the-post (FPTP) can yield some non-intuitive results.

For instance, in 1987 the PCs got zero seats on 29% of the vote. Why? Because the Liberal margin was over 30 percentage points. As you'll see in these other examples, the margin between the parties matters as much as the actual vote percentage.

Just last year in Nova Scotia, at 26 or so percent of the vote each, the NDP and PCs got 7 and 11 seats respectively. The Liberal margin over these tied parties was about 20 points.

In New Brunswick in 1991 CoR and the PCs both got about 20% of the vote, but got 8 and 3 seats. Why such a difference? The CoR vote was more "efficient", that is to say it was concentrated heavily in a band of ridings running from Fredericton to Saint John to Moncton, while the PC vote was spread thinly across the province. CoR got 35% of the vote in the Fredericton region, enough to win plenty of seats as they were only 5 points behind the Liberals. The PCs on the other hand got fairly consistent support of between 19 and 24% in every region, meaning they were always running 20+ points behind.

All that to say, if the NDP gets 20% of the vote and the Liberals and Tories get 35% each the results for the NDP will be much different than if the NDP and PCs both get 20% and the Liberals get 50% (this is closer to what the polls are showing today).

There is however a continuum of seats for the NDP. There are some seats where the NDP has traditionally and consistently done better than elsewhere. And at 20% of the vote in a race with 3 parties at or above that level, it would be virtually impossible for the NDP to fail to win at least a handful of seats. If their vote dips below 15% however, there would be a risk of them being shut out.

I'll break the seats into categories for the NDP's hopes. In most cases I'll be looking at 2006 and 2010 data as a reference point as these are both recent elections and elections where I've been able to transpose the boundaries accurately onto the new map. As appropriate, I'll get into to older data.

Gimmes

This first category of seats are ones where the NDP should pretty well be guaranteed to win in a scenario where they are getting over 20% of the vote. If they don't win these seats, it will be because their vote has collapsed or they've lost due to an unusual vote split or local scandal.

Saint John Harbour

This is the only seat in New Brunswick that has voted NDP more than once. Parts of this current riding were represented by then NDP leader Elizabeth Weir from 1991 to 2005. One might argue that this is not so much an NDP riding as an Elizabeth Weir riding, or perhaps a riding that is sympathetic to electing a party leader. However, that ignores recent results there. Under these boundaries, the NDP would have got 27% of the vote compared to 35% for the PCs and 31% for the Liberals in 2010. This in an election where they got only 10% of the vote provincewide and where their francophone leader was focussed almost exclusively on winning his own seat in northeastern New Brunswick.

Fredericton South

Here is a riding where the NDP came second closest to winning under the new provincial boundaries in 2010. The NDP vote was 21% compared to 32% each for the PCs and the Liberals. Had these boundaries been in place at the 2003 election, the NDP would have won here that year.

Low Hanging Fruit

These are ridings where the NDP has consistently exceeded their provincewide vote in past elections, and where an NDP candidate should be able to win if the provincal vote stays above 20% and the party runs a good campaign in the riding.

Hampton

It is from this area that long-time NDP candidate and party leader George Little hailed. He built a strong following for the NDP here - growing the party to over 30% in 1987 - that hasn't really died out. The NDP strength here is an anomoly compared to the rest of the region. In 2010, the NDP would have taken 19% here compared to 10% provincewide and in 2006 13% compared to 5% provincewide.

Restigouche Chaleur

The NDP would have taken 23% here compared to 10% provincewide in 2010, and 7% compared to 5% in 2006. The fact that they still managed to beat their provincewide vote here in 2006 when they had a unilingual anglophone leader and failed to run a full slate of candidates in the region shows the NDP resiliance. The biggest challenge to the NDP will be that they may face off in a two-way race with the Liberals which would be very difficult for them to win if the Liberals stay 20ish points ahead in the polls. Working in their favour is that long-time incumbent Liberal Roland Haché is retiring.

Saint John East

It was here that the NDP won their second ever seat in a 1984 by-election. NDP strength has remained strong here and they would have taken 19% of the vote under these boundaries in 2010.

Fredericton West-Hanwell

This seat would have given the NDP 17% of the vote in 2010, and 7% in 2006. With party leader Dominic Cardy running here and a focus by the NDP on Fredericton ridings (they have nominated 4 candidates so far, all in the Fredericton area), this riding should be a realistic reach for them.

Call Me, Maybe?

Here are ridings where the NDP has shown some relative strength, but only enough that they would be positioned to win with strong local campaigns, a surge to 30+% in the polls, and/or a closer three-way race between the major parties provincewide keeping the margin between the winner and the NDP to around 10 points.

Bathurst East-Nepisiguit-Saint-Isidore

The NDP would have taken 19% of the vote under these boundaries in 2010, after failing to field a candidate over much of this riding's territory in 2006. The fact that both a PC and Liberal incumbent could create an opportunity to sneak up the middle in a vote split.

Oromocto-Lincoln

The NDP would have taken 12% of the vote here in 2010, and 10% in 2006. The doubling of the provincewide NDP share here in 2006 is owed in part to the fact that party leader Allison Brewer was on the ballot - but that was only for half of the riding and does not fully account for the strong NDP result. The traditional weakness of the Liberals in the Oromocto part of this riding will help the NDP as well.

Saint John Portland

The NDP took 16% in these boundaries in 2010 and 6% in 2006. But under the boundaries of the old Saint John Portland the NDP got over 20% of the vote in 2003 and this riding became more friendly to the NDP under redistribution.

Kent North

NDP support has been inconsistent here but has showed points of strength over the years, most notably in by-elections where the NDP placed second in the old Kent riding in 1998 and 2013. In the last general election, the NDP would have gotten 17% of the vote here though they would have done slightly worse than their provincewide vote in 2006. The key to an NDP victory in this riding seems to be strong Aboriginal turnout. If past NDP candidate Susan Levi-Peters or another prominent member of the Elsipogtog First Nation runs for the NDP they could be a contender. Most of the old Kent riding has gone on to Kent South, however its areas of NDP strength have come here. Interestingly they join with Rogersville-Kouchibougauc a riding which has not generally been strong for the NDP lately, but gave the NDP over 1,000 votes in 1991 and 1999.

Charlotte-Campobello

The NDP would have gotten 14% of the vote here in 2010 and 6% in 2006. My seat model likes this riding for the NDP if the Liberal vote stays below 40% provincewide as the NDP is as likely as the Liberals to benefit from lost PC votes here in that scenario.

Odd Balls

These are ridings which either seem on paper to be good targets for the NDP, but not so much intuitively; or vice versa.

Tracadie-Sheila

NDP leader Roger Duguay got 30% of the vote in the old Tracadie-Sheila riding in 2010. That number is bettered to 37% under these boundaries owing to strong NDP polls that came into the riding from Miramichi Bay and Centre-Peninsule. The NDP did not even field a candidate in Tracadie-Sheila in 2006. My seat model likes this riding for the NDP even when one accounts for the expected "bounce" a leader receives in his riding. This is backed up by the fact that the NDP did better in the polls that came in from neighbouring ridings than they did on average where the leader was actually on the ballot. Coupled with the fact that Roger Duguay did tremendously well in neighbouring Miramichi Bay-Neguac in 2006 (when he was not yet leader), suggests that a strong local candidate could do well here. If the NDP (which had brought in professional operatives from all over for Duguay's 2010 campaign) kept their voter lists and volunteer lists, a strong candidate could be poised to do well here. The NDP campaign would also be helped by the fact that the Liberals have not done well here in two decades and this seat is an unlikely pick up for them.

Miramichi Bay-Neguac

When Roger Duguay ran here in 2006 he bested the provincewide popular vote of his party by 5 times. He had no provincial profile and limited resources. Obviously a strong candidate could do well here. That said, this riding is not as NDP friendly today as it was then. While the NDP took 26% of the vote under the old boundaries in 2006, under these boundaries they would have gotten only 18% (still impressive considering the 5% taken provincewide).

Memramcook-Tantramar

The model doesn't like this riding for the NDP because in recent elections they've done on average only as well here as they did provincewide. Part of that is due to the fact that the Greens ran a strong campaign here in 2010, probably stealing votes from the NDP's potential voter pool. But the strong NDP history here cannot be ignored. In 1982, this was the first seat to elect a New Democrat to the legislature. That was a long-time ago, but the NDP showed continuing signs of strength. In a 1997 by-election, the NDP nearly won taking 30% of the vote to 34% for the PCs and 33% for the Liberals. And they massively beat their provincewide vote in all elections here from 1974 through 2003.

Fredericton North

The model hates this riding because the NDP was crushed here in some parts of the riding in 2006, and didn't have a candidate on the ballot in other parts. In 2010, it managed only to mirror the provincewide average. However, the old Fredericton North gave the NDP 16% of the vote in 2003. That is slightly more impressive if we look at the new boundaries where my estimate gives the NDP 18% of the vote in 2003. Couple this with the NDP focus on Fredericton in general and this riding in particular (candidate Brian Duplessis was the first nominated by the NDP), and this riding is worth watching.

Conclusion

These are 15 ridings in which the NDP has a plausible chance of victory if they hold on to their 20% of the vote. Certainly, there are other opportunities.

For instance, if their campaign focuses on Fredericton and Saint John, their efforts will likely spill into adjacent ridings and ones like Saint John Lancaster and Rothesay could come into play for instance.

If the NDP takes a hard run at northeastern New Brunswick there are also opportunities there. Yvon Godin has proved a New Democrat can win big margins in that region. It is easy to write that off as personal popularity, however there is more to that story. In 1991, the NDP came second in Dalhousie, Restigouche East, Nigadoo-Chaleur and Nepisiguit. In 1995, the NDP came second in Dalhousie and Bathurst, and were just 12 votes short of second in Caraquet. Perhaps a canary in the coal mine that foreshadowed Godin's first breakthrough in 1997? This provincial NDP strength continued into 1999 when they took about a 1,00 votes in each of Nigadoo-Chaleur, Nepigiuit and Centre-Peninsule.

All this is moot however if the NDP strength fails in the heat of an election campaign. We live in interesting political times.

Monday, January 13, 2014

New Brunswick's new ridings

Happy New Year!

This is my favourite kind of year: an election year. It is particularly interesting as New Brunswick will face a new electoral map. This is the first new map since 2006 and only the fourth map since New Brunswick switched to single-member ridings in 1974.

It is my hope to continue with my tradtional subjective analysis of the state of the race in each of the provincial ridings as I did in 2006 and again in 2010. I will supplement that with a more objective poll-based analysis using a seat model I have built for New Brunswick elections. I've been previewing that seat model on Twitter with each opinion poll that has come out in New Brunswick over the past while and will provide more details on it on this blog.

Both the subjective and objective analyses and the explanation of how the latter works will come closer to the election.

First, I'd like to talk about and do some high-level analysis of the new map.

New Brunsiwck has continued its recent trend of fewer districts. After peaking at 58 ridings from 1967-1995, New Brunswick went down to 55 and now to 49 ridings. We now also have a far stricter population variance between ridings, meaning the smallest riding and the largest riding are no more than 10% apart in population. In the past, this range was often around 50% or even more.

Unlike some who criticized the Electoral Boundaries Commission, I happen to think that they have done some excellent work with these boundaries.

Some areas that got considerable criticism which I think were in fact jobs very well done are:

In Fredericton, they have largely corrected the long-time flaw of having the city's immediate suburbs in sprawling rural ridings replacing them instead with urban-suburban ridings which will result in better representation for people.

Gagetown-Petitcodiac is a sprawling riding more than 150 kilometres across, but it consists only of very small municipalities and unincorporated areas with clear communities of interest. For the first time, these voters will have an MLA who can speak to their distinct issues rather than ones whose constituents primarily come from a larger municipality.

Certainly, I would not have drawn all of the boundaries the same and I have a few nitpicks here and there, but this was a very complex task that the commission did very well.

The one place I will criticize the commission is in the naming of some of the districts which is sometimes confusing (Carleton, York, Moncton East are names for previous districts that are re-used in areas bearing little resemblance to their predecessors) and sometimes heavily omit components (Oromotco-Lincoln draws 40% of its population from Fredericton, Charlotte-The Ilses goes far up the coast and includes parts of Saint John, Charlotte-Campobello contains significant parts of York County in the McAdam area). But these are minor details.

Now onto a look at the ridings. I will include a generic statement as to which party the riding generally leans. These statements are not meant to be predictive of the 2014 election, but of each riding's general character. For instance, a riding which would have gone PC in 9 or 10 of the past 10 elections would be called "solid PC", while one then went PC in 5 and Liberal and 5 would be called a toss up.

The guide to these ratings are:
  • solid: this riding will almost always vote for this party;
  • strong: this riding will usually vote for this party but could be lost in a moderate-to-big sweep election like 1982, 1995, 1999 or 2010;
  • lean: this riding will go for this party in a close election or an election won by that party, but may be won by other parties who win with a moderate margin;
  • toss up: this riding will likely go with the overall winner, be decided by local factors, or has an outcome highly dependent on the performance of third parties.
These ratings are generally determined from a two-party lens. Excluding CoR's brief success in 1991, only 3 ridings have gone for other than the two major parties for almost 100 years (and 2 of those did so only once). It is therefore difficult to predict which ridings an NDP or Green candidate might be able to breakthrough into based on past data. With the NDP's standing in the polls, they are likely to be competitive in several seats. In future posts, I'll identify the ridings where the NDP and Greens seem to have a natural support base. In this write up, I will adjust some ridings' ratings to reflect strong third party presence which can make outcomes more unpredictable.

Edmundston-Madawaska Centre

This riding which includes most of the city of Edmundston and some its immediate surroundings and is largely a successor to the old riding of Edmundston-St. Basile from which it draws about 71% of its population, with the balance (29%) coming from Madawaska-les-Lacs. Both of these ridings have been solidly Conservative in recent years, returning PC members since 1995.

This should be a solid riding for the PCs.

Madawska-les-lacs-Edmundston

This is a riding that includes most of Madawaska County, including parts of the City of Edmundston. It excludes the eastern parts of the city, some of Madawska County immediately around the city and the St. Leonard area. It draws about 81% of its population from the old Madawaska-les-Lacs district, 16% from Edmundston-St. Basile and 2% from Restigouche-la-Vallée.

Like Edmundston-Madawaska Centre, this should be a solid riding for the PCs.

Restigouche West

This is one of the ridings whose name iritates me a bit. Restigouche would be more appropriate, as this riding goes clear across the county including all of its western edge, but also most of its eastern edge as well. It excludes the corridor running from the western Campbellton city boundary through Dalhousie to the western Charlo village boundary, and the Belledune area. On the other hand, it includes Campbellton "suburbs" Tide Head and Atholville.

This is in many ways a merger of the 1974-1995 Restigouche West and Restigouche East districts. Restigouche West leaned PC, but went Liberal in good years for them (in 1987, Restigouche West was the closest seat for the PCs in a year in which they were shut out of the legislature). Restigouche East on the other hand was reliably Liberal.

Its immediately predecessors are Restigouche-la-Vallée from which it gets about 39% of its population, Campbellton-Restigouche Centre (37%) and Dalhousie-Restigouche East (24%).

This is seat leans slightly Liberal.

Campbellton-Dalhousie

This riding is by far the smallest geographically in northern New Brunswick. It is about 35km across along Route 11; and generally less than 5km high. It includes the municipalities of Campbellton, Dalhousie, Eel River Crossing and Charlo; as well as a narrow strip of unincorporated areas between Campbellton and Dalhousie.

It draws 54% of its population from Dalhousie-Restigouche East and 46% of its population from Campbellton-Restigouche Centre.

It will likely feature an incumbent-vs.-incumbent battle between incumbent PC MLA Greg Davis (from Cambpellton-Restigouche Centre) and Liberal MLA Donald Arseneault (from Dalhousie-Restigouche East).

This is a seat that leans slightly Liberal.

Restigouche Chaleur

This riding draws primarily from the old Nigadoo-Chaleur riding (about 61% of its population) but also brings in areas from Nepisiguit (26%) and Dalhousie-Restigouche East (13%).

Nigadoo-Chaleur was one of only 10 seats to go Liberal in 1999 and the only one of those that didn't have an incumbent MLA. That as well as more recent trends suggest that on paper, this should be a solid Liberal seat.

However, with the incumbent retiring and the fact that this was one of 9 ridings where the NDP would have taken greater than 15% of the vote in 2010 (23% to be precise), I'll rate it only strong Liberal. If the NDP is to hold its position of 20+% in the polls on election day, this is one of the areas where they'll need to look to convert votes to a seat.

Bathurst West-Beresford

The commission took the decision to split Bathurst in two and incorporate its nearest neighbours in urban-rural hybrids. I applaud this decision as the alternative would have been to set off 20% of Bathurst into a predominately rural riding where it would have been poorly represented. This is a predominately urban riding with 58% of its population coming from the city of Bathurst and another 30% from the urbanized town of Beresford.

This riding is the primary successor to the old Bathurst riding, taking 52% of its population from there. It also takes in about 30% of its population from Nigadoo-Chaleur and 18% from Nepisiguit.

This is a riding that leans slightly Liberal.

Bathurst East-Nepisiguit-Saint-Isidore

The commission took the decision to split Bathurst in two and incorporate its nearest neighbours in urban-rural hybrids. I applaud this decision as the alternative would have been to set off 20% of Bathurst into a predominately rural riding where it would have been poorly represented. This is a mostly rural riding with about 68% of its population coming from outside of the city of Bathurst.

It draws in its population in about a third each from Centre-Peninsule-Saint-Saveur (37%), Nepisiguit (34%) and Bathurst (26%). It also brings in a slight bit of territory from Caraquet.

On paper based on this ridings past voting history, this riding should be solid Liberal. However, we will be seeing an incumbent-vs.-incumbent battle here between PC MLA Ryan Riordon (from Nepisiguit) and Liberal MLA Denis Landry (from Centre-Peninsule-Saint-Saveur) both of whom will be facing a riding in which two-thirds of voters have never voted for them. At the same time, this is another of the potential NDP ridings, where the NDP would have had 19% of the vote in 2010. I will therefore rate this riding strong Liberal.

Caraquet

This is a barely changed riding. It brings 78% of its population from the safe Liberal seat of Caraquet, with 22% of its population coming from the safe Liberal seat of Centre-Peninsule-Saint-Saveur.

This seat is solid Liberal.

Shippagan-Lameque-Miscou

Long the safest Conservative seats in the northeast, this riding has expanded westward to take in some very Liberal friendly polls from Centre-Peninsule-Saint-Savuer.

If the close 2006 election had been fought under these boundaries, the PCs would have eked out a win of just 8 percentage points, compared to the 21 points it was won by under its old boundaries. If the 2010 election were re-run under these boundaries, it would have been won by the PCs by 14.7% - virtually the exact same margin as the provincewide vote.

This seat still leans PC, but it is not the bastion it once was.

Tracadie-Sheila

This riding has been PC for a longer stretch than any other in the province, having gone PC in a 1994 by-election and stayed that way ever since. (One could argue with this, in that Carleton North went PC in a 1993 by-election, and then largely became Carleton in 1995 - however the new Carleton riding bears little resemblance to the old Carelton riding, which is generally a successor to the Woodstock riding. Most of what was Carleton North in 1993, has joined the Carleton-Victoria riding, half of which was Liberal from 1987-2010).

Despite this distinction, it has had a few close calls over the years and has gained some unfriendly territory from Miramichi Bay-Neguac (11% of the new riding's population) and Centre-Peninsule-Saint Saveur (6%). In fact, the PCs ran third in some of these new polls in 2010.

Nevertheless, this riding continues to lean PC.

Miramichi Bay-Neguac

This is another riding with a misleading name. This riding goes far inland from Miramichi Bay, including significant portions of the city of Miramichi and parts north and west thereof. Only 64% of its population comes from the old Miramichi Bay-Neguac riding. A more apt name might simply have been Miramichi-Neguac. Alternatively, it could have been called Northumberland North, with its sister riding, the awkwardly named Southwest Miramichi-Bay du Vin called Northumberland South.

The riding includes the coastal areas running from Neguac down to the city of Miramichi, includes the Douglastown and Nordin parts of that city, then goes north over the city limits inland taking in communities Beaverbook, Sunny Corner, Renous and Red Bank First Nation.

As noted above, about 64% of its population comes from the old Miramichi Bay-Neguac riding, with 29% coming from Miramichi Centre and 7% from Southwest Miramichi.

Based on past election results, this riding that leans Liberal.

Miramichi

This riding consists of almost all of the city of the Miramichi, excluding the Douglastown and Nelson areas. It draws roughly equal parts of its population from the old city-rural hybrid ridings of Miramichi Centre and Miramichi-Bay du Vin.

This riding will be an incumbent-vs.-incumbent affair pitting PC MLA and cabinet minister Robert Trevors (from Miramichi Centre) against Liberal MLA Bill Fraser (from Miramichi-Bay du Vin).

This riding's history suggests it is solid Liberal. In 2010, one of the worst Liberal elections in memory (lowest popular vote for which I can find a record), the Liberals would have won this seat under these boundaries. They would have won by a nearly 2-to-1 margin here in 2006, while tying the provincewide popular vote.

Southwest Miramichi-Bay du Vin

This is one of the largest ridings geographically on the new map stretching the full width of Northumberland County from Green Hill just west of the county line through to Escuminac on the Northumberland coast.

It is principally a successor to the Southwest Miramichi riding (from which it takes about 71% of its population) with the balance coming from the non-city portions of Miramichi-Bay du Vin.

In essence, it is a merger of the 1974-1995 Southwest Miramichi and Bay du Vin districts. Based on its electoral history, this riding is a bit of a toss up. Southwest Miramichi has went PC in 1974, Liberal in 1978, 1982 and 1987, CoR in 1991, Liberal in 1995, PC in 1999, Liberal in 2003 and 2006 and PC again in 2010. The old Bay du Vin ridng went Liberal in 1974 and 1978, but PC in 1982, and Liberal in 1987 and 1991. Its successor Miramichi-Bay du Vin went Liberal in 1995, but PC in 1999 and 2003, before returning red in 2006 and 2010.

Were the 2006 and 2010 elections fought under these boundaries, the PCs would have won in 2010 while exceeding their provincewide popular vote margin, while the Liberals would have done the same in 2006.

This riding is a pure toss up.

Kent North

This riding is largely a successor the old Rogersville-Kouchibouguac district (from which it takes about 78% of its population), with the balance coming from the old Kent district in the Rexton area.

Under these boundaries in 2006, the riding would have gone PC by about 5 points, but this would have been due largely to the popularity of now-Senator Rose-May Poirier. In 2010, under these boundaries, the Liberals would have won by over 10 points despite their worst-ever showing in the provincewide popular vote.

Clearly this is a riding whose voting patterns are tied closely to local candidates and local issues making it a true toss up, though not likely a bellwether, for future elections.

Kent South

Here is another poorly named district. It takes in only about half of the old Kent South district, leaving much confusion for those great many voters who will cast ballots in Shediac Bay-Dieppe or Moncton East in 2014 after having voted in a much different Kent South in 2010.

In fact, the majority of this riding comes from the old Kent district. About 55% of the its population comes from Kent, while only 45% comes from Kent South.

The old Kent riding is one of three current ridings (the others being Bathurst which is being cut in two and Shediac which continues) that has never voted any way but Liberal since single member ridings were created beginning in 1967.

However, due to a solid PC history in the most populous Bouctouche area, it would be fair to call this riding leans PC.

That represents a major change from its namesake (the old Kent South) which was solid Tory country in recent years.

If the trend from the 2013 by-election continues (in which Bouctouche went 2-to-1 for the Liberals), this riding would move into moderately safe Liberal territory.

Shediac Bay-Dieppe

This is a totally new riding bringing together the southern most parts of the old Kent South riding (48% of its population) with significant portions of the two Dieppe ridings (28% from Dieppe Centre-Lewisville and 20% from Memramcook-Lakeville-Dieppe) and a slice of Shediac-Cap-Pele (4%).

Liberal leader Brian Gallant is running here.

Its history suggests that it is a solid Liberal seat, and would have voted Liberal by a 4 point margin despite their worst ever showing in 2010.

Shediac-Beaubassin-Cap-Pelé

Following the taking effect of new ridings in 2014, this will be the last remaining riding that has a perfect Liberal voting record. It is also one of the ridings that has changed the least since single-member ridings became the norm in 1974.

It shed one poll to Shediac Bay-Dieppe and is otherwise unchanged.

This is a solid Liberal riding.

Memramcook-Tantramar

This is probably the most controversial riding on the new map. Francophone groups argued from the outset against moving Memramcook into the predominately anglophone and underpopulated Tantramar riding. For reasons I do not understand, these groups felt it would be acceptable for the equally francophone Cap-Pelé area to be absorbed instead.

In any event, despite initial threats of a lawsuit, ongoing talks with government make it unlikely that these boundaries will change before the 2014 election. That does not rule out a post-election boundary challenge as was successfully the case with the federal Miramichi district after it had been used in the 2004 election but before it could be recontested in 2006.

Despite this change, this riding is still largely the same as the old Tantramar district which accounts for over two-thirds of its population.

Historically, this riding has strongly favoured incumbents. The only incumbent to ever be defeated here was New Democrat Bob Hall in the Liberal sweep of 1987. On other occassions when it has changed hands (1982 and a 1998 by-election) the incumbent had not reoffered.

While the old Tantramar riding has gone strongly Conservative for the past several elections, the incorporation of the Liberal leaning polls from the Memramcook area suggest this riding just leans PC.

This riding is likely to see an incumbent-vs.-incumbent battle between PC MLA Mike Olscamp (from Tantramar) and Liberal MLA Bernard LeBlanc (from Memramcook-Lakeville-Dieppe).

Dieppe

This riding includes much of the city of Dieppe drawing 77% of its population from the old Dieppe Centre-Lewisville riding and 23% from Memramcook-Lakeville-Dieppe. While both predecessor ridings included parts that were not in Dieppe, this riding is purely made up of parts of that city.

Its recent voting history suggests that this riding would been solid Liberal, even voting Liberal were the 2006 election re-run here when the old Dieppe Centre-Lewisville riding went narrowly for the PCs.

Moncton East

This is another poorly named riding. Despite its name, it incorporates only 32% of the old Moncton East riding, whose majority went instead to the new Moncton Centre riding. That said, that small piece of the old Moncton East riding is the plurality population source for this riding which is wholly new, being made up of parts of 6 old ridings, including the old Moncton East (31% of its new population), Moncton Crescent (20%), Memramcook-Lakeville-Dieppe (17%), Kent South (14%), Dieppe Centre-Lewisville (11%) and Petitcodiac (6%).

Its performance in past elections suggests that it should lean Conservative.

Moncton Centre

Despite its name, this riding is primarily a successor to the old Moncton East, from which it draws about 59% of its population; the balance comes from the old Moncton North district.

Though geographically this riding is essentially in the centre of the city of Moncton, its name has drawn some criticism from those who say "centre" implies downtown (or "centre-ville" in French). Moncton's downtown is in the Moncton South riding.

This riding is a true toss up. Were the close 2006 election refought here, my calculations suggest the PCs would have won - by 2 votes! That may be misleading however, as in 2006 the PCs were benefiting from the coattails of Bernard Lord who was re-elected in Moncton East at that time. My calculations suggest the PCs would have won this riding again in 2010 but only by about 1 percentage point, while winning provincewide by about 15 points. One could therefore argue that it leans a bit to the Liberals, but that is a hard argument to make for a riding that would have gone PC in the last 4 consecutive elections.

This riding is likely to feature an incumbent-vs.-incumbent race between PC MLA Marie-Claude Blais (from Moncton North) and Liberal MLA Chris Collins (from Moncton East).

Moncton South

This riding is the most compact of Moncton's electoral districts and includes its downtown. It is largely a continuation of the old Moncton West district (which was once called Moncton South). It draws about 84% of its population from that district, with the balance coming from the old Moncton East (10%) and Moncton Crescent (6%).

In each of the last two elections it would have been won by the PCs by about the exact same margin as they won the provincewide popular vote, making this a true toss up riding.

Moncton Northwest

This riding is in the fastest growing corner of Moncton and is largely a successor to Moncton Crescent. Moncton Crescent was carved off from Petitcodiac in 1995 removing most of the city portions of that district. It was so named, because it was in the shape of a crescent over the top of the city of Moncton. In 2006, it lost its eastern most parts losing much of its crescent shape despite keeping the name. In 2014, it will essentially be just the northwestern third of its old self and is thus quite appropriately named.

It draws 92% of its population from the old Moncton Crescent and 8% from Petitcodiac. The old Moncton Crescent riding was massively overpopulated, such that 92% of the new district, is just 63% of the old, which also gave territory to Moncton East, Southwest and South. So this is not as clear of a continuation as it might sound on its face.

Nonetheless, this like the old Moncton Cresent is the safest conservative seat in the city of Moncton. Rated strongly conservative.

Moncton Southwest

This riding, like the new Moncton East, is a very new district. No former riding brings forward a majority of its population, though a significant plurality (about 48%) comes from Moncton North. The balance comes from Petitcodiac (24%), Moncton West (15%) and Moncton Crescent (12%).

Its history suggest it should be a strong conservative riding.

Riverview

This is one of three ridings whose origins can be traced back to only one predecessor. The old Riverview riding shed 3% of its population to Albert, the rest can be found here.

Like the old Riverview, the new Riverview rates solid Conservative.

Albert

This riding is much the same as the old Albert riding, from which it draws 86% of its population. The balance comes from Petitcodiac (11%) in the Salisbury area, and the territory it gained from Riverview (3%).

Like the old Albert, this riding should be solid Conservative.

Gagetown-Petitcodiac

This is another wholly new riding. It in an interesting riding which stretches nearly 200 kilometres from the Oromocto town line to the Moncton city line. It includes no towns, and a handful of incorporated villages but draws over 80% of its population from unincorporated areas.

Its predecessor ridings are Petitcodiac (45%), Grand Lake-Gagetown (35%), Oromocto (13%), Kings East (5%) and Hampton-Kings (3%).

Though people may have moved in the past 3 or so years, based on the addresses of candidates at the 2010 election - four incumbent MLAs live in this riding! These include Ross Wetmore (from Grand Lake-Gagetown) who is the nominated PC candidate here, Sherry Wilson (who is said to be running in Moncton Southwest), Jim Parrott (who would likely run in Kings Centre if he reoffers), and Jody Carr (who is said to be running in Oromocto-Lincoln).

This seat is solidly Conservative, perhaps their safest seat on the new map.

Sussex-Fundy-St. Martins

Despite the change in name, this is largely the old King East riding. It now stretches down toward the Fundy coast to take in the St. Martins area. It draws its population mostly from Kings East (78%), and also from Hampton-Kings and Saint John-Fundy (11% each).

On paper this is a solid PC riding.

Hampton

This riding is actually quite a bit different from the old Hampton-Kings riding. It has lost most of its geography (including the Kingston Peninsula) and gained parts of Quispamsis and the city of Saint John.

It gets 47% of its population from Hampton-Kings, 18% each from Saint John-Fundy and Saint John East, 12% from Quispamsis and 6% from Rothesay (two rural polls to the southeast of the town).

On paper this leans slightly PC, but it was one of 9 ridings where the NDP got over 15% of the vote in 2010 and the NDP factor makes this riding a bit of a wild card.

Quispamsis

This riding is largely the same as the old Quispamsis (and Kennebecasis before it). Since creation in 1995, this riding has been a perfect bellwether, voting with the government in every election. Its predecessor King West (which included Rothesay and Hampton as well as Quispamsis), also called the winner correctly in elections from its 1974 creation through 1991.

This riding is a toss up.

Rothesay

This riding has brought in some new population from Saint John-Fundy (25% of the new configuration). It goes from having been traditionally a lean PC seat to being a toss up.

Saint John East

This riding is something of a merger of the old Saint John East riding (from which it draws about 59% of its population) and the western part of the old Saint John-Fundy riding (which brings the other 41%).

This riding is one of three in provincial history to have voted NDP (in a 1984 by-election) and the NDP continues to be strong here, getting about 20% of the vote in the last election under these new boundaries.

On paper, it should be a moderately Liberal seat having been won handily by the Liberals in 2006 and lost just narrowly in 2010 beating the provincewide Liberal margin by about 10 percentage points.

Because of the NDP wildcard though, we'll deem it leans Liberal.

Saint John Portland

This riding is changed quite a bit from the old Saint John Portland riding. Its namesake does represent the majority of its population source (56%), but it also draws in population from the old Saint John East (33%), Saint John Harbour (8%) and Rothesay (3%).

This riding comes out as a toss up on paper (it would have gone Liberal under these boundaries in 2006 and was won by the PCs on about the same margin as the provincewide vote in 2010). The tossup factor comes further into play by the fact that this is another riding where the NDP would have scored over 15% in 2010.

Saint John Harbour

This riding is not substantially changed, having just brought in a piece of the old Saint John Portland district to correct for its underpopulation.

Its recent history suggests it should be rated moderately Liberal. However, the strong NDP history here and the fact that this was the NDP's second best riding in 2010 under both old and new boundaries calls that into some question. It will therefore be rated leans Liberal.

Saint John Lancaster

This is the least changed riding in the Saint John region, with 97% of its population coming from the old Saint John Lancaster.

This riding is prone to wild swings, having gone heavily Liberal in 1995 and 2006, heavily PC in 1999 and 2010. It must therefore be rated as a toss up.

Kings Centre

This riding is quite different from anything on the last map, but is somewhat similar to the old Kings Centre riding which existed from 1974 to 1995.

It consists of Grand Bay-Westfield and surrounding areas, then crosses the river and runs up the Kingston peninsula and includes Norton.

The riding draws most of its population from the old Fundy-River Valley district (55%) and also from Hampton-Kings (39%); a small section of population - the Norton area - comes from Kings East (6%).

It rates as moderately Conservative.

Charlotte-The Isles

Here is another poorly named district. Its predecessor got its name because it was part of Charlotte County including two of the three Fundy Isles. Now the riding runs well up the Fundy coast taking in a significant chunk of Saint John County and including a part of the city of Saint John. A better name might have been Charlotte-Fundy.

It takes 71% of its population from the old Charlotte-The Isles ridng and the balance from the old Fundy-River Valley (29%) whose southern/coastal parts it absorbed.

It would have gone Liberal in 2010 - the worst election in Liberal history - meaning it must be called solid Liberal.

Charlotte-Campobello

Again, this riding is poorly named. The riding now stretches well into York County and contains McAdam, which will be the third largest municipality in the district. Charlotte West-McAdam would have been a better name.

It draws 75% of its population from the old Charlotte-Campobello riding, 15% from York and 10% from Charlotte-The Isles.

The riding rates as lean PC, having generally voted just slightly more in favour of the Conservatives than the province as a whole.

Oromocto-Lincoln

Like many seats in the Fredericton region, while I agree with where the lines were drawn, I find the names wanting. This riding draws nearly 40% of its population from the city of Fredericton, yet Fredericton is the only community that goes unnamed here. A better name might have been Oromocto-Fredericton Southeast.

This riding amounts to essentially a merger of the old Fredericton-Lincoln riding and the old Oromocto riding, taking 50% of its population from each. This represents 58% of the population of the old Oromocto riding and 55% of the old Fredericton-Lincoln riding.

The riding rates as lean PC, having generally voted just slightly more in favour of the Conservatives than the province as a whole.

Fredericton-Grand Lake

This riding is probably the best named of the Fredericton districts, but to be clearer on what part of Fredericton is involved a name like Fredericton Northeast-Grand Lake or Marysville-Grand Lake might have been better.

This riding leans Liberal, and would have been lost by the Liberals by only 6 points in 2010, while losing the provincewide vote by about 15 points.

New Maryland-Sunbury

The old riding of New Maryland-Sunbury West was probably the safest PC riding in the capital region. This riding is safer still, taking in the Tory bastion of Geary and area.

It rates as strong Conservative.

Fredericton South

This riding is misleading named. It contains only a small portion what would regularly be considered "Fredericton South" - running only only about 4 km across, while the southside of the city is about 15 km across. A better name might be Fredericton South Centre.

One also needs to be careful not to conflate it with the old Fredericton South riding which existed from 1974 to 2006, and particularly the version of it that existed from 1995 to 2006, as this riding includes much territory that was not in the latter, while excluding much of its territory.

While the old Fredericton South (and its successor Fredericton-Silverwood) had a perfect track record going back to 1974 of voting with the government, one should not expect this riding to be a bellwether.

It is the worst possible riding for the PCs in Fredericton, and it would never have been won by them, not even in their sweep of 1999. The riding would have gone Liberal in 2010, while the Liberals suffered their worst ever defeat. But it would not have been a perfect Liberal riding either. Had this riding existed in 2003, Elizabeth Weir would have been joined by a second New Democrat in the legislature.

On paper this would rate as a strong (but not solid) Liberal riding. However, the strength of both the NDP and the Greens here could make for interesting vote splits. This would have been the Greens best riding in 2010 - nearly 14% of the vote - and it was also a good NDP riding (21%). The PCs could have a hope here in the right election due to vote splits, though their ceiling is likely around 35% of the vote.

Based on these factors, rather than rating it strong Liberal, it will be rated leans Liberal.

Green Party leader David Coon is running in this riding.

Fredericton North

Like Fredericton South, Fredericton North is misleading in its name. Over 42% of northside residents actually live in other ridings. Fredericton North Centre or Nashwaaksis-Devon would have been better names for this riding.

This riding, like Fredericton South, is more Liberal friendly than its predecessor districts due to it being a city centre riding. Its past history suggest it leans Liberal.

Fredericton-York

This riding wins the prize for the worst named riding on the map. This riding includes 12% of the City of Fredericton and 16% of York County. From its name, you might assume that it includes all or most of these areas. We would be just as well served calling this "electoral district 42" as it would be as helpful in its description.

The original proposal of Fredericton-Stanley would be better, but the commission changed it in its amendment to its final report as some of those not in Fredericton and not in Stanley were offended by the name.

Other alternatives might be Fredericton-Nashwaak Valley, Fredericton-York Northeast, Fredericton-Douglas-Stanley, etc.

The riding draws 47% of its population from the old York North district, 38% from the old Fredericton-Nashwaaksis district and 15% from the old Fredericton-Fort Nashwaak district.

The riding leans PC.

Fredericton West-Hanwell

This district isn't terribly named, though Fredericton Southwest-Hanwell might be clearer. Fredericton-Hanwell-Kingsclear might also be a good name for this district.

The district is essentially a merger of the old York district (excluding Harvey and McAdam) and the old Fredericton-Silverwood district (excluding Fredericton's downtown). It draws 56% of its population from York and 44% from Fredericton-Silverwood.

This riding is a toss up on paper. It would have gone Liberal in 2006 but Conservative in 2010. It is also a strong riding for the NDP; they would have gotten 17% of the vote here in 2010.

NDP leader Dominic Cardy is running here.

York

This riding is also named poorly. It includes just 17% of the old riding of York, causing confusion for all concerned. It also includes significant portions of southern Carelton County.

A better name for this riding might have been York West.

This riding is strongly conservative.

Carleton

This riding is named poorly too. It includes just 42% of the old riding of Carleton. A better name would be Carleton South or Woodstock-Hartland.

The riding is solidly conservative. Premier David Alward will be running here.

Carleton-Victoria

This riding is not terribly named, but Carleton North-Victoria would be a clearer description.

This riding is a merger of most of the old Carleton riding (58%) with nearly all of the old Victoria-Tobique riding (86%).

The new riding's population is 57% from Victoria-Tobique and 43% from Carleton.

The riding would have gone Liberal in 2006 but Conservative in 2010, by a much larger margin than provincewide. It is therefore deemed a tossup.

Victoria-la-Vallée

This riding name is a bit odd. It can be traced back to 1995 when the riding of Madawska-la-Vallée was created, getting its name because it contained the part of Madawaska County along the St. John River Valley before that river becomes the border with Maine.

In 2006, when parts of that riding were merged with Restigouche West, the name Madawska-Restigouche was tentatively given to the riding. But because there was a federal riding of the same name that covered much more territory they came up with Restigouche-la-Vallée instead. This made a little bit of sense but less than Madawska-la-Vallée.

Now the new commission has continued this trend, but the problem is that unlike Madawaska and Restigouche counties, Victoria County runs largely along this valley. So, it doesn't really make any sense.

A better name would have been Grand Falls-St. Leonard, Grand Falls-St. Leonard-Drummond or Victoria-Madawaska South.

Interestingly, this riding merges an area that has traditionally been reliable PC territory (St. Leonard) with one that has tradtionally been leaned to the Liberals (Grand Falls).

The result is a riding that is a toss up. It would have gone Liberal by 11 points in 2006 (an election with a virtual tie in the popular vote), but Conservative by 20 points in 2010 (beating their provincewide share by about 5 points).

Overall assessment of the new map

To sum up, we see that this is a pretty fair map with no inherent advantage to either of the main parties. I've labelled 18 seats as favouring the PCs, 18 as favouring the Liberals and 13 as tossups. This equity is mirrored in my seat model. If I plug in for the sake of experimentation an election where the PCs and Liberals each get 50% of the popular vote, the result is 25 seats for the PCs and 24 seats for the Liberals, that slight edge for the PCs is very slight as an election with a popular vote of 50.5% for the Liberals and 49.5% for the PCs models a result of Liberal 25, PC 24.

In the coming weeks, I'll explore areas of potential strength for the NDP and Greens, explain the seat model and begin my subjective "brief look at the ridings" as I did in 2006 and 2010.

Enjoy! I look forward to hearing your thoughts/reactions on Twitter, in the comments section or by email at nbpolitico@gmail.com