New Brunswick

'Astonishingly low' chance of PC win if rejected ballots distributed fairly, court told

The lawyer for former Progressive Conservative candidate Barry Ogden grilled expert witness Peter Loewen on Thursday in a court case that could overturn the election results in the riding of Saint John Harbour.

U of T professor Peter Loewen proposes statistical formulas for distributing rejected ballots

University of Toronto professor Peter Loewen has created two statistical methods for determining how rejected ballots can be counted in cases involving a contested election. (Queen's University)

The lawyer for former Progressive Conservative candidate Barry Ogden grilled expert witness Peter Loewen on Thursday in a court case that could overturn the election results in the riding of Saint John Harbour.

Loewen, a professor of political science, global affairs and public policy at the University of Toronto was a witness for Liberal MLA Gerry Lowe in a court challenge launched by Ogden.

Loewen proposed two formulas Justice Hugh McLellan can use to allocate any ballots he rejects when making his decision on alleged irregularities in the vote.

Lowe beat Ogden by 10 votes in Saint John Harbour in the provincial election last Sept. 24.

Ogden is asking the court to throw out the results and order a new election, alleging 78 instances of a voting irregularity. 

Distributing rejected ballots

Should McLellan determine some or all of the 78 instances qualify as irregular votes he can reject them.

Under normal circumstances, Canadian courts have then removed those votes from the winning candidate's total, a method known as the "magic number."

If the number of rejected votes is greater than the margin between the two leading candidates the election results can be thrown out.

But in a 2012 decision involving the federal riding of Etobicoke Centre, the court admitted the magic number method it is not a good way to settle such matters.

"It inherently favours the challenger," wrote the court in a 4-3 decision. "It assumes that all of the rejected votes were cast for the successful candidate. In reality, this is highly improbable. However, no alternative test has been developed.

"We do not rule out the possibility that another, more realistic method for assessing contested election applications might be adopted by a court in a future case."

In his testimony Loewen offered two alternative methods he claims are fair "transparent and reliable." He dubbed them the "informed method" and the "naive method."  

The informed method takes into account the share of total votes cast for each of the first and second-place candidates and includes that in a formula to determine if the rejected ballots would have made a difference in the outcome.

The naive method takes the total number of rejected ballots and calculates all the potential ways they could have been divided without any reference to the actual vote share.

PC candidate Barry Ogden, left, is challenging the provincial election results in the riding of Saint John Harbour, where Liberal Gerry Lowe won by just 10 votes. (CBC)

For example: In the case of 10 rejected ballots, it would begin with Ogden getting none of the votes and Lowe getting all of them. It would then move on to Ogden getting one and Lowe getting nine. Eventually the combinations would end with Lowe getting no votes and Ogden getting all 10.  ​​

The formula also takes into account the possible ways the votes for the third, fourth and fifth place candidates would factor in.Publish

Loewen then uses a mathematical formula to determine the probability the election result would have changed based on the potential outcomes. He's even developed a calculator to assist in the operation.

Lowe's legal team originally attempted to have the professor qualified as an expert in econometrics, statistics and elections.

But after vigorous objections from Ogden lawyer Matt Letson, who noted Loewen had no graduate degrees in statistics and that his methods were "novel" and untested, McLellan instead certified the professor as an expert in political science.

Under cross-examination Loewen agreed it is possible Ogden could win the election if either of his formulas were used, but the chances are "astonishingly small."

Loewen was recommended to Lowe's legal team by UNB political science professor Joanna Everitt.

She told CBC she is not a Liberal or aligned with any political party.

'Groundbreaking' work

"I think this is an example of how we as academics can actually contribute to discussions around how elections are conducted and managed in Canada," said Everitt.

She said Loewen's methods use statistics to demonstrate the probability of rejected votes falling in favour of one candidate or another.

"This is groundbreaking. It's new, it's precedent setting, and will have significant, I think, impact in future situations where vote counts are being challenged.

Lawyers are scheduled to make final arguments in the case July 30 and 31.

McLellan told the court he expects there will be at least one appeal, meaning the case may not be settled until months down the road.

ABOUT THE AUTHOR

Connell Smith is a reporter with CBC in Saint John. He can be reached at 632-7726 Connell.smith@cbc.ca