The Liberal Leadership delegate results from all 308 ridings are in, but as we noted in our riding-by-riding analysis, the Liberals lost almost half of the ridings (145 ridings) by greater than 15%. Thus, barring a shift of greater than 15 points in the polls (which appears unlikely in the near future), this suggests that the Liberals could win a maximum of 163 ridings (50 safe ridings, 50 that the Liberals won by 15% or less and 63 that the Liberals lost by 15% or less). So, while a leadership candidate may be popular in Alberta or rural Quebec, there is little chance that popularity will translate into electoral success in those ridings in the near future (since the Liberals lost these ridings by an average of 40 points). A candidate’s popularity in the 163 winnable ridings, by contrast, could make the difference between whether the Conservatives or Liberals win the next election. Here are the number of delegates for each candidate in those 163 winnable ridings — it is important to emphasize that these numbers are objective; they merely represent the difference between competitive and out-of-reach ridings (here defined by a plurality of 15%, although a plurality of 10% or 20% shows similar proportions).

	IGNATIEFF	RAE	KENNEDY	DION	VOLPE	DRYDEN	BRISON	MHF	UNDEC.
WINNABLE	26.1%	21.4%	20.5%	13.5%	7.1%	5.3%	2.8%	0.6%	2.7%
OUT-OF-REACH	35.2%	18.7%	12.8%	21.5%	2.7%	2.9%	3.2%	1.0%	1.9%
DIFFERENCE	-9.1%	+2.7%	+7.7%	-8.0%	+4.4%	+2.4%	-0.4%	-0.4%	+0.8%

ANALYSIS

Michael Ignatieff
It is clear that many of Ignatieff’s delegates are in ridings that are out-of-reach for the Liberals. In fact, only about half of Ignatieff’s delegates are from winnable ridings (617 of 1160). This can be attributed to this high number of delegates from Alberta and rural Quebec. Still, given his consistency across the country, he has the most delegates among winnable ridings with 26.1%, but that is just 4.7% ahead of Bob Rae and just 5.6% ahead of Gerard Kennedy. In general, Ignatieff has significantly more support in out-of-reach ridings than in winnable ridings (35.9% vs. 26.1%).

Bob Rae
More of Bob Rae’s delegates come from winnable ridings than Ignatieff’s — almost two-thirds of Rae’s delegates are from winnable ridings (506 of 794). In general, Rae has slightly more support in winnable ridings than in out-of-reach ridings (21.4% vs. 18.7%), but the difference is relatively small, suggesting that Rae’s delegates are reasonably balanced.

Gerard Kennedy
Kennedy significantly improves his position among winnable ridings — 71% of his delegates are from winnable ridings (484 of 681). Although Kennedy scored well among out-of-reach Alberta ridings, he was shut out from out-of-reach Quebec ridings. By constrast, he performed well in Ontario’s 905 and Southwestern regions. In general, Kennedy has significantly more support in winnable ridings than in out-of-reach ridings (20.5% vs. 12.8%) — a difference of 60%.

Stephane Dion
Dion scored significantly worse among winnable ridings than his overall totals. In fact, Dion received slightly more delegates from out-of-reach ridings than from winnable ridings (331 vs. 320), overall a roughly 50-50 split. Clearly, this is due to Dion’s success in Alberta and rural Quebec, and his weaker totals in Ontario and Atlantic Canada. In general, Dion has signficantly more support in out-of-reach ridings than in winnable ridings (21.5% vs. 13.5%) — a difference of 60%.

CONCLUSION
It is clear that the ability to win the next election is becoming an increasingly important criteria for the Liberals (although most recognize it is also necessary to reform the party). The above analysis sheds some light on which candidates are most popular among Liberals members. The low-turnout and low proportion of the population who are Liberal members makes inference to the overall electorate impossible. At best, the delegate results illustrate which candidate might best rally Liberal supporters. And the ability to get out the vote could make a difference. There are obvious consequences — it is not a stretch of the imagination to think that rallying the base in winnable ridings will be more difficult if their preferred candidate is not chosen. This analysis is clearly good news for Bob Rae and Gerard Kennedy, as it demonstrates that they are much closer to Michael Ignatieff in ridings where it counts. By contrast, this analysis is bad news for Ignatieff, as it demonstrates his lead is much narrower in ridings where it counts. Likewise, it is bad news for Stephane Dion, as it demonstrates he is much weaker among winnable ridings than his overall numbers suggest. So on the basis of electability, three candidates appear to have an advantage: Michael Ignatieff, Bob Rae and Gerard Kennedy, who are with 5.6% of one another.