Vote Splitting is a Myth

exercise-bike-race

Here are some facts for you:

  • George Bush beat Al Gore in the 2000 presidential race because Ralph Nader was siphoning off Democrat votes.
  • The BC Liberals [a right leaning party] keep winning because of vote splitting on the left.
  • The NDP won in Alberta because of vote splitting on the right.
  • The federal Liberals won repeatedly during the 1990s because of vote splitting on the right.
  • The federal Conservatives are in power today because of vote splitting on the left.

Do they all make sense, or just the ones where your favoured party lost? Here’s the kicker: none of them are true. Now you are thinking that this can’t be, especially for the elections that didn’t go the way you hoped, but reality tells a different story.

Vote splitting doesn’t change the result for two reasons:

  1. Those who are concerned about vote spitting already vote strategically, and
  2. The political spectrum is not really a spectrum in the proper sense of the word. By that I mean that two ideologies thought to be at the same location on the political spectrum can actually be at odds with each other politically.

Let’s take the example of the union left and the environmentalist left. Both are important players within the New Democratic Party (NDP), but appeasing one group can alienate the other. British Columbia NDP leader Adrian Dix found this out the hard way during the 2013 provincial election when – in an effort to shore up environmentalist votes that could drift to the Greens  – he alienated the union vote by declaring the Trans-mountain pipeline dead no matter what. The union vote abandoned ship in droves, which reversed the fortunes of the NDP.

Many union jobs are in forestry, mining, and oil and gas, and though the rank and file support unionization and good wages for the working man, they also support maximizing resource extraction (with more union jobs). This is why many voters who work in the resource industries go back and forth between the NDP and the Conservatives, but can’t stand the Liberals (whom they consider the worst of both worlds).

The Liberals are thought of as more centrist than the NDP, but on some issues they are further to left. One example is gun control where the NDP in rural Canada finds itself more aligned with the Conservatives than the Liberals. This is largely why the Liberals win almost no seats in rural Canada while the NDP does quite well.

In addition to NDPers who drift right, many Liberals also prefer the Conservatives over the party to their left. There’s the free enterprise coalition in British Columbia made up of Conservatives and right leaning Liberals that (in the words of W.A.C. Bennett)  “vote to keep the socialist hordes [NDP] behind the gates of Hell where they belong.”

To be sure, there is a large voting block on the left that sees the Conservatives as the enemy, and votes strategically when possible. This movement largely assumes that everyone who isn’t voting Conservative would vote for the Devil himself if it meant getting rid of the Conservatives. These folks feel quite jaded with the political system when they see that a Conservative candidate won with only 49% of the vote while the combined “progressive vote” (whatever that means) was 51%. This frustration stems from the assumption that one plus one equals two. In other words, everyone who votes Liberal, NDP, or Green self-aligns as “progressive voter” and a champion of “social justice.” This is of course totally false, and in fact, A + B can sometimes be less than A! Or if you prefer numbers to letters, 1+2<2.

A+B<A? Sometimes it is. For example, the electoral riding of Skeena-Bulkley Valley, currently held by the NDP’s Nathan Cullen, was not an NDP riding when the right was divided.

Skeena-Bulkley Valley

Figure 1: Voting history of the Skeen-Bulkley Valley riding.

In 1987 the right-wing populist Reform party formed in opposition to the governing Progressive Conservatives. The movement grew rapidly, and by their second election in 1993, they became the second largest national party in Canada.

Skeena-Bulkley Valley2

Figure 2: 43+3=34

Since the Reform party was to the political right of the Progressive Conservatives, the pundits assumed that they garnered all their support from the right side of the political spectrum. This assumption was false.

In reality, much of the Reform party’s success lay with western alienation and other populist ideas that span the political spectrum. Standing up for western rights in the face of a large  and powerful government really struck a chord with the union voters that traditionally supported the NDP.

The two right-wing parties merged prior to the 2004 election and this quashed the populism of Reform. Many of those on the left who had supported Reform moved back to their natural home with the NDP.

Figure 2 highlights what happened after the two right-wing parties merged. As competing entities, they had managed 46% of the vote between them (43% from Reform), but after the merger they could only muster 34%. This in turn relinquishing the seat back to the NDP.

Second C

Figure 3: Second choice for Wildrose supporters.

And how about the NDP’s victory in Alberta? Surely, the impossible could only be achieved by vote splitting on the right. That’s a reasonable assumption, but it’s wrong. The voters were that upset that they threw out the PCs all on their own.

Had the Wildrose not existed, the NDP would have not only still won, and by a wider margin. We know this because of polling data asking the right-wing Wildrose and the center-right Progressive Conservative (PC) voters for their second choice. The results revealed that they did not go for each other in high enough numbers to prevent an NDP victory. In fact, Wildrose supporters chose the NDP in higher numbers than the PCs, which should not be too surprising given the disillusionment voters felt toward the government.

Second Choice r

Figure 4: Distributing Wildrose votes.

Only 21% of Wildrose supporters chose the PCs as their second choice. Redistributing these votes per Figure 3 reveals that the NDP would have won an even stronger majority, taking 68 seats – 14 more than they did in the 2015 election.

And you are thinking that the PCs were extremely unpopular, so second place PC data should be looked at instead. Well, the result is similar. PC supporters were more likely to pick the Wildrose as their second choice, but still,the NDP comes out on top with a solid majority at 56 seats, two more than they won in the real world.

On the federal level, a Liberal blog called Blunt Objects has analyzed what would happen in the event of a merger between the Liberals and NDP, and the results are not any better for the merged parties. This blog goes into private mode during election campaigns so I can’t link to the analysis, but suffice to say that 1 plus 1 does not equal 2, especially in British Columbia. This is because many NDP and Liberal supports would rather vote Conservative than the other party. Blunt Objects concluded from the 2011 polling data that a merger would have boosted the Conservative’s majority.

Ultimately, governments defeat themselves (just ask the Alberta PCs), and vote splitting has nothing to do with it. The Federal Conservatives will eventually lose, just as their predecessors lost over the Sponsorship Scandal, and that’s a good thing. As they say, politicians and diapers need to be changed frequently, and for the same reason. Voters are perfectly okay with lying from within their own camp, and will tolerate a tremendous amount of corruption before they switch their vote, but eventually the scales tip. In British Columbia the NDP lost power, not because of vote splitting on the left, but because enough voters wanted the NDP out of office. The rise in the BC Green vote in 2001 and the rise in Wildrose vote in 2015 had everything to do with an electorate wanting change so badly that they didn’t care who won so long as it wasn’t the NDP or the PCs.

Blaming a party’s lose on vote splitting is a lame excuse. It’s not better than professional athlete blaming vote splitting for the reason he didn’t win the MVP, or an actor who didn’t win an Oscar doing the same. We see many examples of the “vote splitting” card being played when it’s simply not an issue. Terry Glavin has a few egregious examples such as the NDP claiming that voting Green will cause the Conservative to win, knowing full well that the race was between the Greens and NDP.

Liberal and NDP supporters talk about a merger, but the Liberal leadership resists this idea for good reason. For the Liberal party, a merger with the NDP would mean the loss of some support to the Conservatives, and worse, among supporters with the deepest pockets. For the NDP, a merger makes more sense, which is why they are more supportive of the idea.

There’s another political party who would like to see fewer parties on the left, and that is the Conservatives. Stephen Harper’s goal was to put the Liberal party out of business, and leave only two viable national parties – the NDP and the Conservatives. It should not be lost on the average voter that both the Liberals and the Conservatives agree that fewer parties on the left benefits the Conservatives.

One more point about vote splitting and strategic voting. As was stated earlier, people already vote strategically. What I mean by that is that the voters who are concerned about vote splitting already vote strategically.

Let’s give an example. On a scale of 10 to -10 (where -10 means you’ll jump off a bridge if they get elected), Sally ranks the parties as:

  • NDP = 6
  • Liberal = 5
  • Conservative = -9.

Johnny ranks the parties as:

  • NDP = 6
  • Liberal = 5
  • Conservatives = 5

Sandy ranks the parties as:

  • NDP = 6
  • Conservative = 5.0000001
  • Liberal =4.99999999999

Garth ranks the parties as follows:

  • Conservative = -7
  • Liberal = -8
  • NDP = -9

Only Sally is concerned about vote splitting, and thus will vote for the candidate best able to defeat the Conservative candidate, while Johnny, Sandy, and Garth will vote for their preferred candidate. None of these three are splitting the vote because they don’t rank one party much worse than the others. If you have a sizable portion of NDP and Green supporters that see little difference between the Liberals and Conservatives, are they really splitting the vote? Obviously they are not.

Vote splitting implies that voters with one particular ideology are torn between two parties, while another party wins because they are the only party representing another ideology. It’s like having three groups of people fighting over what to eat at Tim Horton’s — Boston cream donuts, Maple glazed donuts, or bran muffins. If the doughnut lovers are divided equally, and the bran muffin lovers win with 35% of the vote, then you might be tempted to conclude that vote splitting caused the muffins to win. This may be true at the doughnut level (though we should remember that some people don’t like chocolate), but at the political level it doesn’t work like this. We have daily polls weeks in advance of voting, giving ample time for those who would prefer doughnuts of any kind to muffins to vote strategically. Maple glazed fans who refuse to cross over to the more popular doughnut don’t see much difference between a Boston cream doughnut and a bran muffin, therefore, they are not splitting the vote.

Here’s further thought. The madder the electorate, the more they are going to vote strategically to remove the party in power. Therefore, the fact that a lot of people are voting strategically is not a sign of a flawed system, but a sign of a political party that, in the eyes of the electorate, is corrupt, inept, and dangerous.

In summary, strategic voting exists, but vote splitting does not.

Now go out and vote.

Advertisements
This entry was posted in politics. Bookmark the permalink.

One Response to Vote Splitting is a Myth

  1. Pingback: The Trump Landslide Myth | Questioning The Data

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s