Logical myths of vote-splitting
Two plus two equals three — three-and-a-half if all goes well
Second of four parts
Much was said about vote splitting during the run-up to Wildrose and PC members’ endorsement of the United Conservative Party. Since then, more speculation about how many former PC and Wildrose voters the UCP will attract has ensued. Unfortunately, the simple math most often used — adding up the votes two parties received in a previous election or in recent public opinion polls to total what a new party will receive — does not do justice to the complexities of voting behaviour and the impact of the electoral system when translating votes into seats.
Initially, we must remember that Canadians are known as vote switchers. This results in all parties losing some of their supporters between elections, typically at least one-fifth and sometimes half or more. As such, all parties engage in an ongoing process of electoral coalition rebuilding, constantly seeking to maintain as many supporters as possible while reaching out to new recruits to replace those they will lose between elections and increase their overall support.
Federal data illustrates this phenomenon. During the last three elections in which the Progressive Conservatives competed, the PCs retained only 22 per cent (1993), 51 per cent (1997) and 44 per cent (2000) of the voters who supported them in the previous election. In other words, in increasing their total share of votes between 1993 and 1997 they had to make up for losing half of their past supporters before getting around to increasing their total.
During that same period, Reform/Alliance was much more efficient at maintaining its base, consistently retaining 80 per cent of its previous voters. Yet their task was similar. In increasing its total vote share by six per cent in 2000 the Alliance had to replace one-fifth of its former votes prior to securing those overall gains.
The situation is even more complicated for merged parties like Alberta’s united Conservatives. Federal data indicates that in its first election in 2004 the merged Conservative party retained 88 per cent of former Alliance voters and 68 per cent of former PC voters, resulting in slightly less than 30 per cent of the vote, approximately three-quarters of that received by the legacy parties. In fact, it took the CPC three elections to equal the totals previously achieved by the PCs and Alliance.
Alberta polling data indicates a similar pattern emerging in united Conservative support. Prior to the merger, Mainstreet Research registered combined Wildrose and PC support at 57 per cent. Postmerger numbers indicate the UCP has 43 per cent. A substantial base from which to build, but again only about three-quarters of previous legacy party support. Further, the number of undecided voters has surged from 15 per cent to 27 per cent. Given the relative stability in their competitors’ numbers, most of that increase appears to be former legacy party supporters who have yet to be convinced of the virtue of the new party and are still shopping.
As misleading as it may be, the simple two-plustwo-equals-four logic is persuasive and will therefore likely continue to be used to oversell the benefits of ending vote splitting. Albeit unnecessarily because a merged party does not need to win over as many voters as its legacy parties to be more successful than were the legacy parties.
Federal conservative numbers are again instructive. Despite winning eight per cent fewer votes than the Alliance-PCs total in the 2000 election, in 2004 the CPC won 21 more seats than its two legacy parties. In 2006 the CPC still won a smaller percentage of the total than its legacy parties, but won 124 seats compared to the 2000 Alliance-PC total of 78. A substantial improvement despite not yet reaching the combined total of the legacy parties.
So it is fair to say that ending vote splitting helps merged parties like the UCP, for a variety of reasons, not least of which is the simple elimination of one choice for voters. But knowing the UCP will both gain and lose some of the legacy parties’ former supporters need not compel proponents to oversell the benefits of ending vote splitting any more than it should embolden detractors who are doing little more than pointing out the obvious in suggesting the new party will not garner all former supporters of its legacy parties.
The ultimate fate of the UCP will be determined by how well its leadership maintains and perpetually renews its support base.
Much of that will be importantly influenced by factors beyond its control such as the state of the economy and its competitors two years down the road. Much, however, is under its own control. Only the UCP will be responsible for its leadership, policy positioning and strategic decisions. These will ultimately determine how successful any overall UCP voter mobilization strategy will be.
We will turn our attention to some of those matters next week.
Faron Ellis is Research Chair, Citizen Society Research Lab, Lethbridge College. George Rigaux is a Lethbridge banker, Reform-CPC organizer and campaign manager.