As touched on in my first post on the Reform Act, some critics of the bill argue that formal rules establishing a procedure by which a party caucus could initiate a vote of confidence in, followed by the possible removal of, its leader aren’t necessary since caucuses already have that power. Alice Funke, for example, writes:
there is nothing in the law currently preventing party caucuses from doing this very thing now, and indeed they have done so frequently in our current system: Joe Clark was pushed into a leadership review, Michel Gauthier was pushed out as leader by the Bloc Québécois caucus, a good part of Stockwell Day’s caucus left him and the Canadian Alliance and joined the remainder of the Tories instead. And a significant group of Paul Martin backers were hatching plots to oust Jean Chrétien as Liberal leader and Prime Minister.
Today on Twitter, Ms. Funke linked to this article which describes how the British Columbia Social Credit Party caucus forced their leader (and at the time Premier) Bill Vander Zalm out in 1991.
It is certainly true that even without formal guidelines or rules in place, a caucus can exert enough pressure to force a leader to resign, or at least agree to a leadership review at a party conference, but the process can be a lengthy, messy and often very public one which can end up being quite detrimental to the party in the long run. One only needs to think of the Chrétien-Martin divide which hurt the Liberal Party of Canada long after the fact, or the Blair-Brown divide which similarly plagued the UK Labour Party.
The Vander Zalm case is, in fact, a good example of why specific rules would be a good idea. Because the party itself did not have a process in place to allow the caucus to trigger an internal leadership vote of confidence, the party members had to resort to using a parliamentary procedure to achieve what they could not: they planned to table a motion of non-confidence in their own government. Ultimately, the motion wasn’t needed as Vander Zalm was found guilty of violating conflict guidelines and stepped down voluntarily.
There are a number of problems in using this example to prove that caucuses don’t really need formal rules to trigger for possible leadership change. First of all, simply put, I can’t help but think that this is a misuse of the Confidence Convention. The SoCreds were unhappy with their leader, not the fact that their party formed the Government. Confidence of the House is given to a Government; who heads that government is an internal matter for the governing party to decide. Related to this, this option is one that could only be used by a party that formed the government. Opposition parties cannot move want of confidence motions in the House against themselves or their own leader – they can only move want of confidence motions against a sitting government. Consequently, an opposition party unhappy with its leader can’t go this route.
Another problem is that it isn’t necessarily guaranteed to result in a change of leadership. As I explained in my previous post, the tradition in Canada for governments which lose confidence votes is not to resign, but to seek dissolution and trigger a new election. In the article, it appears as if the Social Credit caucus kept the Lt.-Governor informed of what was transpiring, explaining that they planned to “withdraw majority support from Vander Zalm and delegate it to another of their number.” And apparently the Lt.-Governor agreed that he would ask Vander Zalm to resign rather than agree to a dissolution. But what if the Lt.-Governor had not agreed to listen to the caucus? What if the premier had decided to pre-empt his caucus and seek a dissolution and new election? To put it simply, the party should not have had to go this route; if they were unhappy with their leader, they should have simply been able to resolve that internally without resorting to moving a want of confidence motion in the Government.
I am not a constitutional expert, but a lot of this sounds like involving the Crown in the internal machinations of a political party and that makes me somewhat uncomfortable. As per House of Commons Procedure and Practice, 2nd ed., “no act of the monarch (or Governor General as the monarch’s representative) is carried out without the formal advice and consent of the Prime Minister and Cabinet.” There is no mention of carrying out the advice of a party caucus. I will leave that issue to persons better qualified to comment on.
Many critics of the Reform Act worry that an empowered caucus will lead to chaos, with party leaders being shown the door on a regular basis. Is this necessarily what happens?
In Politics at the Centre: The Selection and Removal of Party Leaders in the Anglo Parliamentary Democracies, William Cross and André Blais compare leadership procedures in Australia, Canada, Ireland, New Zealand and the United Kingdom. It is important to know that of these five countries, Canada is the outlier, the only one where political parties have no specific entrenched procedures available to caucuses to challenge their leader. Chapter 5 looks at how and why party leaders’ tenure in office ends. This normally occurs in one of three ways: the leader resigns (either voluntarily, or under pressure), they are forced from office, or either they, or their party, die. Cross and Blais examined the departures of 110 leaders between January 1965 and January 2008 and found that most (76%, or 84 of 110) resign, and the majority of those who resign (59, or 54%), did so under pressure. It probably goes without saying that most of these leaders who resigned under pressure did so to avoid being forced out. (p. 97)
What we are most interested in, however is how many leaders were actually forced from office. Over the 43-year period studied by Cross and Blais, only 17 leaders in those five countries were forced out: New Zealand 5, Australia 7, the UK 3, Ireland 1, and Canada 1. The Canadian case – John Diefenbaker, is the only one removed by “a process involving the extra-parliamentary party.” (p. 106) Additionally, it is important to note that most of these forced leadership changes occurred among opposition parties. During that same time frame, only 3 sitting Prime Ministers were forced out by their own caucus – Margaret Thatcher (1979-1991) in the UK, and Bob Hawke (1983-1991) and John Gorton (1968-71) in Australia. (p. 98) Thatcher had been in power since 1979, but by late 1990, the Conservatives had been trailing Labour for 18 months in the polls. These same polls showed that a change in leadership would give the Tories a lead over Labour. Bob Hawke’s popularity had been in decline from the late 80s, and while he led Labor to a narrow re-election in 1990, his party lost faith in his ability to counter the resurgent Liberal Party. John Gorton simply proved to be a poor choice for leader, and in his first general election as Prime Minister (1969), saw the Coalition’s 45-seat majority over Labor that he had inherited reduced to only a 7-seat majority. He was forced out as leader of the Liberal Party not long afterwards.
These examples demonstrate two important points. First, the caucuses of parties that are in government aren’t likely to force a leader (and Prime Minister) out if the party is doing well in the polls. The three Prime Ministers forced to resign by their caucuses were forced out due to declining polls or poor election results. The second point is that it is parties in opposition which are more likely to force a leader out, and these are the parties which would not be able to use the want of confidence approach to leadership change described in the Vander Zalm piece or as postulated by Dale Smith.
Of course, Cross and Blais’s research does not take into account the Rudd-Gillard-Rudd leadership spills of 2010-2013, but even by Australian practice as described in Politics at the Centre, that chapter was an anomaly. And as I stated in my earlier post, Labor has now changed its rules governing challenging a sitting leader; consequently, we will never seen anything like that occur again.
I admit that I still don’t quite understand the arguments of those who insist that formal rules for triggering a leadership change aren’t at all needed. I think Canadian political parties would benefit from having formal rules in place. Is the process outlined in Chong’s Reform Act the best way to proceed, perhaps not. I am not comfortable with these rules being incorporated into the Elections Act; I think they should be adopted by parties and included in party constitutions. And I certainly fail to understand how anyone can insist that the confidence convention is a viable option for parties to effect leadership change. It isn’t, and even if it were, it would only work for a party in government. Parties in opposition would be left with no clear options. Hopefully this short international context I’ve provided will help calm a few fears. The process isn’t abused by political parties. It doesn’t result in political chaos. If anything, it might avoid it.