Labour: Britain’s Unity Party

Alex Salmond, Scottish First Minister.

Scotland will hold a referendum on independence in the autumn of 2014, at the prompting of the Scottish Assembly’s Scottish National Party (SNP) majority government. At the moment it appears that the pro-independence camp will lose. Polls place support for Scottish independence roughly in the 30% to 40% range. Another indicator of the limited appeal of independence in Scotland can be found in a recent Guardian/ICM poll on the British monarchy. While 36% of Scots said Britain would be better off without the Royals, a surprising 50% of Scots said the country was better off with them. One would assume that if at least 50% of Scots feel the nation is better off with the monarchy that at least 50% of Scots are not interested in breaking away from the United Kingdom. While support for out-right independence remains below the winning post, opposition to the SNP in general has also declined notably over the last decade, as indicated by the share of Scots who claim they would never vote SNP.

What this suggests is a trend in Scotland that in many ways mirrors the last fifty years of history in Quebec: a rise in support for greater regional autonomy within a united nation. The two extremes in Scotland of strict unitary unionism and independence are not the preferred option of most Scots, just as top-down federalism and independence is a binary that most Quebecois reject. This is why when I as a Canadian look across The Pond at the constitutional wrangles and nationalist debates within Britain, I merely see a British version of the constitutional sorting out that dominated Canadian domestic politics from the early 1970s until the late 1990s. Two referendums on Quebec independence, a failed constitutional revamp attempt and a nationwide referendum on another constitutional framework occurred in the interval, and Canada emerged as a more decentralized, asymmetrical federal state, and paradoxically enough, a more united one, at least insofar as the question of unity faded as a national issue. What is merely happening in Britain is the gradual transformation of the country into a federal state.

The transformation of Britain into a united yet federal state however could be derailed by the presence of Conservative-Liberal Democrat coalition government. The Tories, never a party that warmed up to the idea of Scottish and Welsh Assemblies, maintain that the question of Scottish independence is a matter that only the Parliament at Westminster can deal with, and that such an issue is beyond the purview of the Scottish Assembly (also known as Holyrod). By insisting that the matter of independence cannot be decided within Scotland itself, or rather by the Scottish Assembly, risks generating support for independence by implying that Scots do not have the freedom to make such a vital decision themselves. In many ways the impact would resemble the rise of support for independence in Quebec that occurred in the past whenever Ottawa insisted that it had the final say on the province’s future within Canada. The strict unionist message of the Tories (and by extension their Lib Dem partners) is where the Labour Party has an opportunity to sell itself, both within the Celtic fringe and outside of it, as the party of British national unity.

A not particularly well-placed election sign for Labour in Scotland during the 2010 general election.

The Labour Party initiated both the unsuccessful 1979 referendum on Scottish devolution (albeit with some quirky stipulations) and the successful 1997 referendum on the same matter. Therefore, Labour already has credibility as the party that is more willing to bring about a federal Britain. Moreover, the history of post-war British politics has shown Labour increasingly become the party of Scotland, while the Tories have become a fringe force to the north of Hadrian’s Wall.

As the chart above shows, during the first twenty years following World War II, both the Conservatives and Labour commanded equivalent levels of loyalty over Scottish voters. Beginning in the early 1970s, however, the Tories began to experience a long-decline of support in Scotland, only surpassing 30% of the vote once in the last 35 years and dropping to below 20% since 1997. In the last three British general elections, the Conservatives came in fourth place in Scotland in terms of party support. Labour support, meanwhile, has proven to be rather resilient in Scotland despite the decline of the monolithic two-party system that dominated British politics between 1945 and 1970. Labour support in Scotland has hovered in the 40% range for the last 40 years, despite some awful nationwide results for Labour. Indeed, in the 2010 election, Labour support actually rose in Scotland, despite a nationwide decline in support for Labour to below 30%.

The share of seats won by each party in Scotland since 1945 further exaggerates the Tory decline and Labour bedrock in the country. Since 1959 the Labour Party has held at least half of Scotland’s Parliamentary seats, and at least two-thirds of Scotland’s seats in every election since 1987. The Tories, meanwhile, have seen their seat haul from Scotland gradually decline to almost nothing. No Tories were elected in Scotland in 1997 and only one seat has been won by the party there in every election since. Even if the seats held by the Tories’ Lib Dem coalition partners are added to the total, the current Cameron-Clegg coalition hold only one-fifth of the seats in Scotland. Only once since 1945 has a British government (John Major’s government between 1992 and 1997) been more poorly represented in Scotland. Further, with the Lib Dems down in the 7% range in the polls in Scotland, the prospect of Lib Dem MPs being elected in Scotland in the next election are rather slim.

Simply put, the Conservatives lack the support in Scotland to effectively represent the country and lack the policy positions, not just on devolution but also on a host of economic and social matters, to win over the hearts of Scottish voters. Labour can present itself to Scotland as the pan-British party that best reflects Scottish values at Westminster and can present itself to the rest of Britain as the party that can guarantee that the vast majority of Scottish constitutional ambitions are satisfied within a united Britain, thus preventing an emotionally charged and messy divorce. Labour will have plenty of other angels from which to attack the Cameron-Clegg coalition in the next general election, but the argument that only Labour can secure a united Britain will be a novel and useful additional selling point. In doing so, there will be a further parallel with Canada’s unity crisis of the past. Canada’s Liberals used to be the party of national unity, yet their outdated concept of federalism could not accommodate modern Quebec, and instead it fell in the 1980s to Mulroney’s Tories, and today to Mulcair’s New Democrats, to offer a pan-Canadian option that meets Quebec’s constitutional needs. In offering itself as the mid-wife to a united and federal Britain, the Labour Party will obtain yet another hammer with which to bash the Conservatives.

Quebec, 2012: Fury in the Streets, Fluctuation in the Polls

Anti-tuition increase march, Montreal, March 22, 2012.

Even before the anti-tuition increase-inspired student strike came to dominate Quebec’s public discourse, voters in La Belle Province were demonstrating a willingness to engage in erratic and drastic shifts in their party preferences. The student strike has not halted this process but has simply added a new dimension to the battle for support among Quebec’s political parties among an electorate whose loyalties are in constant flux.

Before turning to the wild shifts in party support in Quebec in the last six months, let’s look at the political mood around the student strike itself. It is clear that both sides in this dispute are engaged in a grand tug-of-war for overall public support and sympathy. The Charest government and the striking students entered into this showdown with equal levels of support according to Forum Research, who recorded 38% support for each side in late March. Up until last week, the students were on the losing end of the tug-of-war, seeing their support drop to 33% by May 15, while the government’s support increased to 45% over the same period. The decline in support for the students was likely due to distaste for the actions of the more militant segment of the movement. It was becoming clear that few of the student strikers were attending classes in public relations beforehand. Things took a sudden turn last week when the Charest government passed new emergency strike legislation. The students saw their support uptick a bit to 36% while the government’s support dropped to 40%, all within a period of 48 hours. The PR battle between the students and the governments has once again become an open contest, reversing what looked like a slow but inevitable public opinion loss for the students. What is notable is that for the last two months the share of Quebecers who support neither side has remained fixed at around 20%. It is likely that some of those in this camp eventually shifted to supporting one side while partisans for one side shifted to the neither camp, but what is apparent is that both sides in this dispute have yet to significantly dent the perception among many that both sides are behaving badly.

How is the student strike impacting the battle for support among Quebec’s political parties? It is clear that the Charest government’s stance, thus far, has not halted the slow but steady increase in Liberal fortunes since late 2011. Monthly polling averages for December 2011 pegged Liberal support at 25.0%, while polling averages for May put the Liberals at 32.0%. The Parti Quebecois (PQ), meanwhile, have been on a rollercoaster ride of support in the last six months, jumping from third place and just 21.0% in December to first place and 36% in March and down to second place and 30.5% in May. The PQ’s declarations of support for the students’ stance is certainly firming up the party’s trade unionist, academic and leftist base, reflected in the simultaneous decline of the socialist Quebec Solidaire party (likely due to the PQ absorbing some of that party’s softer support), yet it also appears to have cost them their comfortable lead, putting them in a one-on-one fight with the Liberals. The Coalition Avenir Québec (“Coalition for Quebec’s Future” in English), or CAQ, has meanwhile fallen from the giddy heights the party experienced upon its formation in the autumn of 2011. I expected that the CAQ would experience a drop in support as its policy positions became more apparent, simply because a lot of CAQ support in the polls was people imposing their own political desires upon a blank slate of a new party. The absorption of the old ADQ into the CAQ fold, while removing a rival for the CAQ on its right flank, has obviously damaged the CAQ brand, at least in the short-run. The CAQ is now firmly identified as Quebec centre-right party, and the new party represents to many a mere transfer of ADQ wine into CAQ bottles. While this association certainly provides the CAQ with a political niche and a ready base of support, it also has dampened the enthusiasm of some voters who were hoping for a different type of party to emerge. The CAQ has collapsed from a monthly polling average high of 38% and a solid lead in December to third place and just over 20% in April. May has at least seen a stop to the CAQ’s bleeding, but the student strike could do further damage. Voters may feel that the new party, despite possessing some talent from previous governments, may not be up to the hard challenges that are to be faced.

The ever-shifting political landscape in Quebec is even more starkly reflected in the seat projections based on the last six months worth of polls. Using a Quebec seat projection simulator developed by Bryan Breguet of Too Close To Call, we see that the monthly polling averages, when converted into seats, produce a wildly different set of outcomes within a relatively short time-frame.

Polling average-to-seat conversions for December 2011 put the CAQ in comfortable majority government territory, with 76 seats (63 are needed for a majority). The Liberals form the Official Opposition with 28 seats while the PQ win 19. Fast forward two months and the PQ wins a bare majority of 66 seats, while the Liberals still maintain Official Opposition status at 40 seats and the CAQ plummets to third place and 18 seats, a figure that continues to decline in projections until May. As of May, we see the slow Liberal climb progressing and the PQ dipping, such that the seat projection for the month gives the Liberals a slim minority of 57 seats against 56 for the PQ. The upshot of this exercise is that attempting to anticipate how Quebecers will cast their votes anytime before the next provincial election is actually called is a fool’s game, and speculations on how the seat count will play out in an erratic three-party race is even more difficult in the absence of an official election campaign.

The chart below offers a quick glimpse into the regional variations in party support in Quebec, based on polling averages for the month of May. The regional averages for May 2012 are compared to the regional results in the December 2008 provincial election.

Averages derived from polls conducted by Leger Marketing (May 2), CROP (May 3), and Forum Research (May 15, 17). CAQ figures for 2008 are for the ADQ.

We see in these comparisons that the Liberals and PQ are down in every part of Quebec since the last election, though the Liberal drop is greater, particularly in the Montreal area. The CAQ results (which are compared to the ADQ’s 2008 results, as the CAQ is the effective heir of the ADQ) are mixed, with notable increases in the Montreal area and non-big city Quebec yet a slight decline in the Quebec City region. The CAQ’s Montreal numbers bode well for potential gains in the off-island suburbs. Quebec Solidaire, meanwhile, sees it support up across the province, and their rise in Montreal also opens up the possibility of an extra seat or two from them on Montreal Island.

The final impact of the Quebec student strike is impossible to gauge at this point in time, both for the actors directly involved and for the Quebec political landscape as a whole. As evidenced above, Quebec was already in an atmosphere of great political flux, and it would be surprising if this strike does not further this mood. As it stands right now, the new dynamic in the contest between Quebec’s political parties might be summarized as such: that the Liberals and PQ, both failing to maintain a level of support on par with their 2008 results for most of the previous year, are in an odd race to see who can lose fewer of their original supporters.

Translating Polls Into Results

Anyone who follows polls know that they come with a host of caveats, namely the sample size and the resulting margin of error of a particular survey. Margins of error can explain in part the gap between polls and election results, yet there are other reasons why polls can sometimes be less than accurate in anticipating the actual popular vote shares in an election.

The polling conducted during the last week of the 2011 Canadian election and the results of that election serve as an interesting illustration of how polls can do everything right yet still need to be qualified as a means of gauging actual support levels. I calculated the national and regional averages of the final polls conducted by polling firms during the last week of the campaign, based on these figures provided at threehundredeight.com. I averaged polls that used the same regional divisions for the sake of uniformity, so the averages I came up with used the final polls conducted by EKOS, Harris Decima, Forum Research, Abacus, Angus Reid, Ipsos-Reid and Leger Marketing. I compared the polling averages to the actual national and regional results, the outcome of which is presented in the charts below.

Figures are percentages. A positive figure means that a party’s regional election result was higher than the polling average. A negative figure means the a party’s election regional result was lower than the polling average.

The first chart shows that on the national level, the polls understated eventual Conservative support while overstated to varying degree the support levels of other parties. On the regional level the polling vs. result gaps are even more pronounced. The polls significantly overstated NDP support in Atlantic Canada while understated Conservative support in the region. The Liberals saw their poll to result fortunes decline as one moved westward, with the polls understating their support in Atlantic Canada yet overstating it west of Ontario. The Conservatives were understated everywhere, while the Greens were overstated everywhere. The NDP had mixed results, with the polls being generous to the party in Atlantic Canada and British Columbia yet still not capturing the full scale of the “Orange Wave” in Quebec and being more-or-less spot-on in the Prairies and Alberta. So, why was this the case? Afterall, these polls were all conducted using proper scientific polling methods and surveyed adequate sample sizes from a representative demographic base. The likely answer is no fault of the pollsters but the gap between the general public as a whole and the share of the public that is likely to vote.

Polls capture public opinion from a sample that seeks to match the demographic breakdown of the country, getting responses from a sample that matches the distribution of people by sex, age group, income and educational levels and region. In other words, polls show what the result of an election would roughly be if everyone voted. However, everyone does not vote, and this is reflected the gap between polls and results. It is known that voting rates are greater among older voting groups and are dismally low among young voters. A detailed look at polls that provide the age range and income breakdown of the responses will show that Conservative Party support tends to be higher among older voters while Green support tends to be highest among younger voters. In other words, Conservative supporters are more likely to actually cast a ballot compared to Green voters, and this can explain the nationwide understatement of Conservative support in 2011 and the nationwide overstatement of Green support. The NDP and Liberal results are more mixed because their support bases are not quite as easily defined along age lines. The NDP used to be notably more popular among young voters than older voters, yet recent polls show more voters in older age ranges moving to the NDP than in elections past, which bodes well for their translating of polls into results. This trend also might explain why the gap between national polling averages and election results has narrowed for the NDP since 2004. The gaps between party polling levels and election results are also to a degree the result of differing Get-Out-The-Vote (GOTV) efforts. The Conservatives have the most sophisticated GOTV machine in the country, followed by the NDP, while the Greens’ GOTV machine is barely functional.

How does the political scene in Canada look when the information above is applied to the current crop of polls? Below are two charts. The first illustrates the regional polling averages for April 2012, as compiled by threehundredeight.com. The second chart is what those regional polling averages are when altered according to the regional poll-to-result gaps registered in the 2011 election.

As one can see, there are some noteworthy differences between the raw polling data and the altered polling data. The strong NDP lead in Atlantic Canada declines and the region becomes a straight-up three-way race. The Bloc Quebecois suffers further losses, while the NDP loses its lead in British Columbia and is a bit dampened in Ontario. The Liberals, meanwhile, enjoy a small boost in the Atlantic region yet fail to mount any sort of rebound in the West. The Conservatives enhance their lead in Ontario and the Prairies and take over in B.C., while the Greens’ level of support slides from low to lower across the board. The differences between these two sets of regional polling numbers can have a notable impact on seat projections. When regional result-based seat projections are done for each sets of data, the outcomes are as such:

Using the original polling numbers, the NDP actually manages to eke out a one-seat victory over the Conservatives, while the Liberals make a decent rebound, largely due to dropping Tory support in Ontario. When the altered polling results are converted into seats, the Conservatives instead win a comfortable minority and are clearly ahead of the NDP, who still manage to grow their caucus yet fail to win a first shot at forming a government. The Liberal rebound, meanwhile, is also muted, again due to failing to knock of Tories in Ontario.

One should not take the information I’ve offered here as hard truths that can be applied with certainty to future polls. These merely point to past tendencies that should be considered when assessing any polls. Improved or degraded GOTV efforts by different parties between now and 2015 will make a difference, as will changing levels of party support among different age groups. Indeed, one could attempt to convert polls into more accurate reflections of public opinion via a calculation that takes party support by age group and alters it to reflect voter turnout rates. The primary tendency that can be confidently taken from this study is that polls will likely understate Conservative support and overstate Green support, and only obvious changes in either the age base of each party’s support, or increased youth voter turnout rates, will change this tendency.

When examining the altered polling averages from April, there are a few assumptions we can make about how the political scene has changed in the year since the last election. The Conservatives are clearly down across the country, while the NDP is definitely up, most notably in the Atlantic region, the Prairies and B.C. The Liberals have appeared to, for the time being, halt their decade-long process of bleeding, though have not reversed it, which puts them in better shape than the still-bleeding Bloc Quebecois. The Greens, meanwhile, continue to inhabit a weird zone between the serious parties and the fringe. A year since the last election, there are reasons for the Conservatives to worry yet not panic, for the New Democrats to celebrate yet not become over-confident, and for the Liberals to consider their future in the context of what their role and purpose as a third party is to be.

Danielle Smith’s New Challenge: Keeping Wildrose Together

There’s no doubting that the Wildrose Party and its leader Danielle Smith had a disappointing night on April 23rd as the results of the Alberta provincial election came in. The polls and the pundits (including this lowly political junkie) were predicting that the 41-year reign of the PC Party would end and that Smith would become Premier. There is however some consolation that Wildrose can draw from the election outcome.

First, their showing was only disappointing by the standard of the polls during the campaign itself. At the beginning of 2012 and throughout 2011 the polls suggested that predicting a result for Wildrose like the one that they achieved was an act of high optimism. Indeed, none of the polls conducted between March 2010 and the dropping of the writ in 2012 pegged Wildrose support at the level they actually won. That Wildrose managed to win just over 34% of the vote is the second consolation that they can draw from the election. The share of the vote won by Wildrose is the third best achieved by an opposition party during the entire PC Party reign. Only the Liberals in 1993 and Social Credit in the PC’s first electoral win in 1971 secured higher levels of support. The Wildrose caucus is not of a bad size either by Alberta standards, proportionally the largest Official Opposition in a decade. Moreover, the margin of loss for many Wildrose candidates was rather slim, ensuring that the party can translate a mere handful of PC converts into significant seat gains. If only one in ten of the voters who opted for the PCs in 2012 were to switch to Wildrose, the party would win 38 seats compared to 40 for the PCs. Thirdly, becoming Official Opposition instead of government may be a blessing in disguise for Smith and here new enlarged caucus, for it will potentially give her party time to grow and mature as a government-in-waiting and provide some much needed experience to a largely untested collection of MLAs. To have had so many politically inexperienced candidates suddenly thrust into government could very well have ended in disaster for Wildrose. That being said, things could very well still end in disaster for the party as an Official Opposition, and the potential internal party conflicts that may arise will very much test the party’s ability to be seen as either a government-in-waiting or a ramshackle flash in the pan.

WCC = Western Canada Concept, an Alberta separatist party.

All political parties (apart from a few hyper-ideological extremist sects) are effectively coalitions, yet the two major components of the coalition that is the Wildrose Party have the strong potential to become bitter foes quickly. The party consists essentially of two wings: an urban/suburban libertarian wing and a rural/small city socially conservative wing. Both wings sit firmly on the right on economic matters, yet it is on social policy questions and the role of the state in regulating individual morality that these two wings are at extreme odds. Social conservatives believe in what is called a “perfectionist” notion of individual lifestyle, or rather that there is a defined set of lifestyle choices, personal behaviours and family/living arrangements that are sound and that ought to be followed. These perfectionist notions go beyond the accepted basics of not killing or raping people or committing arson. They extend to consensual sexual relations between adults and family types. Hence the social conservative obsession with homosexuality, contraception, abortion, single-parent and same-sex parent households and traditional gender roles. Promoting, or even outright enforcing, a perfectionist ideal of such individual behaviour is seen as a legitimate role for the state. Libertarians, on the other hand, believe in a wide degree of individual freedom free from state or communal coercion. All personal choices and interactions are legitimate as long as they are contractual. According to libertarian philosophy, the individual is only obligated to not infringe upon the freedom of action of others, and matters of one’s consensual sexual activity, living or marriage arrangements, reproductive choices, expression via words or visuals and recreational habits are of no concern to the state and not a legitimate arena for legislation. Social conservatives and libertarians may agree that healthcare delivery ought to be either private or charitable, yet they would have very different ideas about the right of a woman to access a free market-provided abortion. The same conflicts would arise on matters like drug and alcohol use, school curriculum and gay rights.

The challenge for Danielle Smith will be deciding how she as a libertarian manages the social conservative wing of her party. It is evident that comments by some socially conservative candidates during the campaign cost the Wildrose Party significant support, especially in increasingly cosmopolitan Calgary, and the libertarian wing of the party will be hard-pressed to not point fingers at the social conservatives for costing them power. Smith will also face problems with the fact that most of her new caucus originate from rural Alberta, and thus are more likely to be on the social conservative wing of the party and demand policy debates and express positions that further alienate urban and younger voters, not to mention the party’s libertarian wing. It is already evident that Smith intends to soften or eliminate some of the party’s policies that dissuade urban and younger voters, and as such social conservative positions, and air-time for socially conservative ideas, would be certain to go. Smith may very well be forced into a situation where she must reassert the Wildrose Party as a strictly libertarian party and either silence or expel the social conservatives in the hopes of making the party more electable. However, the social conservatives may make the decision for her. If they are silenced or ignored too much, will they leave on their own and form a strictly social conservative party, one that would be dead in the cities but have a base in rural Alberta? The possibility is heightened by the overly rural composition of the Wildrose caucus, and disgruntled social conservative Wildrose MLAs could very well end up forming their own splinter party in the Legislature. Ultimately each wing will have to determine at what point they cannot live with the other, and at what point the benefits of working together within a single party are outweighed by the negatives.

The potential conflict brewing within the Wildrose Party offers an interesting illustration of the broad conservative coalition that has emerged in North America since the early 1970s. Radical free-market anti-Keynesian economics became a force just as politically active social conservatism (in the form of the Religious Right) did so as well, and as such they both gravitated to the traditional centre-right parties as they offered a much more hospitable vehicle for their separate causes than the parties of the centre and the centre-left. It is for this reason that contemporary conservatism is a collection of such curious bedfellows. In most centre-right parties, the social conservatives and the libertarians are both minority forces, and as such their policy and worldview disagreements are less likely to come into direct conflict, as they are both battling a conservative middle. There is no such conservative middle in the Wildrose Party. As such, these two extreme ends of the modern conservative coalition directly face eachother, and the future of Wildrose will serve as an interesting case-study of how well these two extremes can co-exist.

Political Party Word Association

Political parties have an interesting task on par with any company that is selling a product: they must both work at constructing and protecting a “brand” identity, or rather a set of feelings, values and ideas that the voter or consumer associates with the product. Parties offer voters a package of policies, personalities, values and philosophies, yet when operating in the open market of democratic politics, parties also benefit or suffer from automatic associations that voters make in their minds towards a particular party. Parties will have certain brand strengths and weaknesses and must work to enhance the former and combat the latter. Nanos conducted a rather interesting poll in mid-April 2012 that asked Canadians what words they associated with the five political parties currently represented in Parliament. While the poll specifically asked people to give one word that would describe each party as a “person,” the results offer insights into the brand of each political party. In other words, what automatic associations do voters make towards a particular party?

The answers were open-ended, and as such a lot of answers had to be grouped into general responses, yet the detailed results give a good indication of the most common word associations the respondents made about different parties. What made me curious was to examine which words or terms that respondents gave were positive, negative and subjective. Below is a list of terms referred to in the data and how I categorized them.

It was hard to regard terms like “good”, “competent”, “caring” and “intelligent” as anything other than positive, and to regard terms like “bad”, “untrustworthy”, “arrogant”, and “selfish” as anything other than negative. I put “Environment/Eco-friendly” and “Progressive” into the positive camp because while everyone has different ideas of what environmental policies should be enacted and different ideas of what constitutes progressive, it is generally agreed now that the a degree of environmentally-friendly policy is positive and that being progressive, in the vague sense of moving forward, is also positive. In other words, they’re like “democracy” or “human rights.” There may be disagreements about the meaning, but the terms are accepted goods.

Most the subjective terms were ideological descriptions. Afterall, the term “conservative” is a badge of honour for someone on the right and an insult for those on the left. “Centrist/Middle of the Road” is to some a sign of balanced thinking and to others a sing of weak compromise. “Socialist” is an almost pointless term in the early 21st century, as it can describe the whole spectrum from Bolshevik to Barack Obama, depending on who you ask. Moreover, a lot of people call themselves “socialist” when they really mean “social democrat.” In other words, their definition is much less severe than the meaning I would ascribe to the term, which would be a heavily nationalized command economy. I even put “Hippie/Radical” in the subjective column because while most people use these terms derisively, there are those who take great pride in such monikers. “Aggressive” and “Idealistic” were included as subjective terms as well because the terms can be either insults or praises depending on who one asks. “Aggressive” conjures up the idea of “mean” for some and “strong-willed” for others, and “idealistic” can be regarded as both “principled dreamer” and “naive know-nothing.”

After dividing up the terms given as response into the three categories, I added up the percentage frequencies that positive, negative and subjective terms were used per party. These were the results:

Percentages do not add up to 100% due to some responses being unable to categorize, such as "None" and "Other."

As we can see, the NDP garnered the highest share of positive responses, the Bloc Quebecois the highest share of negative responses, while voters were least ambivalent or subjective in their feelings on the Conservatives and especially the Liberals. What is interesting regarding the terms associated with the Bloc Quebecois is that the responses regarding them are from Quebec voters only. It would not be of much bother to the Bloc if non-Quebec voters held such a low regard for the party, yet the fact that their fellow Quebecois seem to express such a negative set of terms in relation to the party does not bode well for their party brand and any hope of a comeback.

To get a sense of the relative overall health of a party’s brand, I subtracted the percentage of responses that were negative from the percentage of responses that were positive.

As we can see, the Conservatives, Liberals and especially the Bloc Quebecois have an overall negative party brand, as the negative terms about them outweigh the positive ones. The Greens and especially the NDP have overall positive party brands. Things becomes particularly good for the NDP and particularly bad for the Liberals and Bloc Quebecois when one makes the same calculation yet treats the subjective terms as entirely positive or negative.

If we regard the subjective terms given by respondents as being entirely negative (in other words, calling the NDP “Socialist” as a slur as opposed to a compliment), we see that the NDP is the only party whose share of terms that are positive still out-weigh the terms that are negative and subjective. Every other party falls into a negative net score in such a calculation.

If one treats the share of subjective terms for each party as wholly positive, things improve for the Conservatives and Greens. Their net scores become positive. This is particularly important for the Conservatives in the branding of their party. If they can convince voters that subjective terms associated with them like “conservative” and “aggressive” are virtues, they have a shot at reversing an otherwise overall negative party brand. The potential of party brand rehabilitation is less of an option for the Liberals. Of the four nation-wide parties, only the Liberals still suffer from and overall negative party brand if all of the subjective terms respondents associated with them are regarded as positive. In other words, the Liberal Party brand is the most injured and battered of the four nation-wide parties. Not exactly the best place to start from if the Liberals are seeking to halt and reverse their decline. The Bloc Quebecois figures, meanwhile, need little elaboration on how dire they are for them.

Observing these results, we can conclude that the NDP has the strongest and healthiest party brand, that the Liberal brand is greatly suffering and the Bloc Quebecois brand is moreorless mortally wounded, and that the Greens and Conservatives have party brands of mixed value, largely dependent upon who one asks. The areas in which parties are regarded positively and negatively and the overall status of their party brands offers some insights into the strengths that parties will use to promote themselves, the weaknesses of their own that they will try to hide or mend, and the weaknesses of opponents that they will try to use against them, between now and the next federal election.

Alberta Prediction Review: What Happened?

One of the most entertaining aspects of democracy is the capacity of voters to surprise, to throw out of whack all of the analysis and polling and predictions that are made beforehand of their intentions. This is what happened on election night in Alberta, where the Progressive Conservatives (PCs) managed to secure another majority government and the Wildrose surge faded.

Before looking at why the voters confounded the pollsters and the predictions, let’s evaluate my own prediction. I got the general outcome totally off, with the predicted first place and second place party reversed and predicted third and fourth place party reversed as well. On predicting individual seats, I got 48 correct out of 87, or 55.2%. Not exactly great, though this sort of score appears to have been universal among those who do this stuff. By my count I performed just slightly worse than 308.com, who got 51 of 87 seats correctly predicted (58.6%), and performed better than bluntobjects, who correctly called 42 out of 87 ridings (48.3%). Needless to say, we all didn’t do too well.

So, what explains the gap between what the polls said during the last week of the campaign and what the results were? It is apparent that two major things occurred that were related; the first is that Wildrose’s campaign became burdened by controversies regarding candidates and policies in the final week, and the second is that the result was the vast majority of undecided voters (along with tepid Wildrose supporters) going to the PCs. Some media commentators on election night kept talking about strategic voting by Liberals, yet this explanation isn’t sufficient. As the polls during the campaign show, the Liberal vote had already declined significantly from their 2008 result when the writ was dropped, and their final result (9.9%) was barely below their last week polling average (11.0%).

In other words, the PCs had pretty much already won over a lot of Liberal supporters before the election was even called. What really cost Wildrose the election was the controversies surrounding some of their candidates and policies and how this shifted undecided voters. The trend away from Wildrose and towards the PCs was evident during the last weekend, and internal PC Party polling on the weekend gave them a four-point lead. What simply happened was that media polling ceased just as that final shift was taking place. The only poll to capture the beginnings of this shift was conducted by Forum Research on the last day of the campaign, which showed a Wildrose lead of only two points over the PCs. This represented a Wildrose drop of three points and a PC increase of four points since the previous day. Had the trend been extrapolated one day forward (and thus to election day), the PCs would have had 40% to Wildrose’s 35%. In other words, this final Forum Research poll was the only public clue that things were shifting very quickly in the final 48 hours of the campaign.

It was stated that as many as 20% of the electorate were undecided on the weekend, and if this was the case, by my calculations about 85% of undecided voters opted for the PCs once in the voting station. There was little data beforehand on where undecided voters were leaning. The most recent indicator I could find was a Leger Marketing poll conducted on April 17 that featured a breakdown of the parties that undecideds were leaning towards. Of those who gave an answer, about 36% were leaning towards the PCs compared to 25% for Wildrose. If the results of this poll regarding where undecided voters were leaning were applied to the last week’s worth of polling averages, and an undecided rate was 20% was assumed, a prediction of the final vote tally would still have put Wildrose at about 37% to the PCs at 34%. My own prediction used the Leger data and assumed a 10% undecided rate going into election day. Even if I was armed with that final Forum Research poll, my method of predicting the final popular vote total would have been 38% for Wildrose and 34% for the PCs. A more narrow gap than predicted but the seat count would certainly have still been a Wildrose win, majority or minority. It certainly wouldn’t have been the 9% popular vote gap and three-and-half times seat difference that actually occurred. Scanning the data that is available, it is evident that a last minute stampede of undecided voters towards the PC Party en masse is what secured their victory.

Final Alberta Election Prediction: The Tory Dynasty Dies

I have completed my final riding-by-riding prediction of the 2012 Alberta provincial election, and if it goes the way I’ve anticipated it will on Monday evening, 41 years of Progressive Conservative government will come to an end. The upstart libertarian-hard conservative hybrid known as the Wildrose Party will secure a majority government, winning 55 out of 87 seats. The PCs will be reduced to 26 seats, half of them in Edmonton. Alison Redford, however, will at least be victorious in her own riding of Calgary-Elbow. The NDP will double their caucus to a grand total of four, while the Liberals will be nearly obliterated, relegated to a rump of two seats in Edmonton. In terms of popular vote, I’m predicting that Wildrose will garner 39.4% of the vote. The PCs will drop to 33.0% of the vote, down from 52.7% in the 2008 election. The Liberals will come in third in popular vote despite running fourth in terms of seats, securing 13.0%, while the NDP will win 11.1% of the vote. The Alberta Party will win 2.1%, while the remainder (Evergreen, Separation, Communist, various Independents) will collectively win 1.4%. Premier-elect Danielle Smith will be tasked with forming a new government with an almost universally inexperienced caucus, which has proven to be the norm in Alberta political history. In the three changes of government in Alberta’s history, the new governing party won zero seats in the previous election on two occasions, while Peter Lougheed’s PCs won only six of 65 seats in the election before their provincial victory.

A seat-by-seat breakdown of the prediction is as follows:

Here is my guide for election night viewing, with seats that should offer some interesting races and potentially surprising outcomes:

Potential NDP Pick-Ups: Apart from the two extra seats I’ve given to the NDP (Edmonton-Beverly-Clareview and Edmonton-Calder), there is the potential for NDP pick-ups, assuming the vote splits all work in their favour, in Edmonton-Centre, Edmonton-Glenora, Edmonton-Manning, Edmonton-Goldbar and Edmonton-Ellerslie. There has also been some talk of NDP potential in Lethbridge, though I’d be shocked if an actual NDP win happened here, as opposed to a strong performance.

Potential Liberal Redoubts: The Liberals could very well be wiped out on election night, and this election is very much a full-scale defensive operation for the party. Don’t expect the Liberals to win any seats they don’t already hold, and don’t expect much outside of Edmonton. The one possible exception I’ll be watching is Calgary-Mountain View. This Liberal riding, held by David Swann, posted the highest Liberal vote in Alberta in the 2008 election, slightly more than the Edmonton-Riverview riding of then-leader Kevin Taft. If the Liberals have any shot of holding a seat outside of Edmonton, this is it. However, with the Liberals polling in the 10%-15% range in Calgary, down from 33.9% in 2008, holding this riding may be just too much of a long shot.

A Declaration of Independence in Sherwood Park?: The suburban Edmonton riding of Sherwood Park could offer a unique surprise in the form of Independent candidate James Ford. Ford lost the federal Conservative nomination in Edmonton-Sherwood Park in 2008 and ran as an Independent, winning 32.5% of the vote. He ran again in 2011, winning 29.5%. These are phenomenal figures for an Independent candidate, and one is forced to wonder how well he will do on the provincial scene in a smaller geographical area (one that he presumably is better-known in). Sherwood Park could very well be a three-way right-wing fight.

The Controversial Wildrose Duo: I’m certain lots of political observers will be paying particular attention to the ridings of Edmonton-Southwest and Calgary-Greenway, where the Wildrose Party are running Allan Hunsperger and Ron Leech respectively. While I am certain you already know why these tow candidates are controversial, here is a refresher on Hunsperger and Leech. I predict that Hunsperger will lose yet Leech will win, as he secured almost 26% of the vote in Calgary-Montrose in 2008 as an Independent. In other words, he seems to have a base.

All in all, Alberta’s election night on Monday promises to be a dramatic event, even if the PCs somehow manage to hang on. Should the prediction I have put forward turn out to be what happens, the possible ripple effects upon the province’s political scene, including the very survival of the Alberta PCs and Liberals, will be significant indeed.

Oooooh, the colours! My final prediction mapped out. Green = Wildrose, Blue = PC, Red = Liberal, Orange = NDP.