Skip to content
Join our Newsletter

Les Leyne: NDP win won’t come from a computer

The thinking in some B.C. New Democrat circles now goes something like this: If only there was a software package the party could buy to win an election. Download BeatChristy 1.1 (for Mac and PC) from Amazon, say, and install it at headquarters.

The thinking in some B.C. New Democrat circles now goes something like this: If only there was a software package the party could buy to win an election.

Download BeatChristy 1.1 (for Mac and PC) from Amazon, say, and install it at headquarters. The program would analyze voter information, format the right micro-targeted messages and “devise a data-driven action program,” to quote from the experts. Follow the prompts and the program would deliver an NDP government as smoothly as Google delivers search results.

There’s no such thing. But some NDP members on Monday unveiled what looks like an attempt to build something similar.

I don’t think it’s going to do the trick. But you can’t fault the enthusiasm of the backers of the Winch Institute, which opened its online doors for business this week.

It’s going to sponsor research in “the science of voter contact and data management” by gathering like-minded statisticians and mathematicians to model systems that pull the vote and win elections.

Andrew Mercier, a young former candidate who lost his race in May, said the concept sprang from admiration for the data management and outreach done by the Democratic Party in winning the U.S. 2012 presidential election.

NDP MLA Bruce Ralston talked with others about modelling that work and the Winch Institute, named after former NDP leader Harold Winch, was hatched.

The Democratic Party’s effort has become the stuff of legend. After John Kerry’s defeat in 2008, the party made over its backshop voter analysis and get-out-the-vote operations. The Winch Institute has a vision of “serving as the analytics department for the progressive movement” in Canada.

“Through rigorous application of the most current techniques in analytics, the Winch Data Initiative will give the progressive movement a decisive advantage in future elections.”

There’s a lot of emphasis on the success of the Democratic Party model. But that discounts another key factor — Barack Obama. He had at least as much to do with their two wins as the backshop did. And one of the big differences between the Kerry loss and Obama’s wins wasn’t just a computer-program upgrade — they came up with a better candidate, as well.

Another point about the new analytics shop is that the NDP’s opponents are doing exactly the same thing.

A B.C. Liberal strategist wouldn’t go into details, but left the impression the Liberals will not be left behind when it comes to staying current with political data analysis. He said the topic has a certain allure as it is viewed as a mysterious black box that can work miracles. But it’s not that complicated. The NDP lost in May because people didn’t like their policies, he said.

They can tinker with the mechanical works, but they were driving the wrong car in the wrong direction, “with probably the wrong driver, as well.”

The NDP already has a similar operation established within the party. Former campaign manager Brian Topp described it in his post-mortem of the loss that was released last month.

He said they implemented a number of interesting new campaign tools in 2012, “including a micro-targeting system that helped us predict, quite accurately, which voters might tell us they supported us and would contribute financially to our campaign.

“This micro-targeting system was part of a larger data-management system that allowed us to track, record and preserve substantially all of our local and central voter identification work, for the first time.”

But Topp concluded it didn’t work.

“It is more than a little depressing to compare the ‘marks’ we recorded in target ridings through these well-designed and impressive systems with the actual turnout,” he wrote.

Maybe that’s an argument for the Winch Institute to step in and design a better system. But maybe it shows there’s too much emphasis on analytic projections that can’t be trusted.

Just So You Know: The party executive is now recommending May 25, 2014, as the date to pick a new leader. The idea has to be approved by the provincial council. Thirty-six of the 133 council members are on the executive, so the vote will indicate how restive the party is when it comes to taking direction.