Canadians still think it's a good idea to negotiate with Afghanistan's Taliban insurgents as a way to end the violence there, a poll finds.

In The Strategic Counsel poll conducted for CTV and The Globe and Mail, there was almost two-to-one support for the notion:

  • Net good idea: 63 per cent
  • Net bad idea: 32 per cent

The proportion of respondents saying it was a bad idea dropped by four percentage points when the same question was asked in October.

"In a way, it's a very Canadian thing to believe that nothing can't be solved by sitting across a table and talking," Peter Donolo of The Strategic Counsel told on Sunday.

However, Canadians might also think the mission is a morass, with no real end point in sight, he said.

Donolo said 57 per cent of Conservative Party members supported the idea of negotiations.

When NDP Leader Jack Layton called for peace talks with the Taliban last fall, Foreign Affairs Minister Peter MacKay later called the approach "naive." Some wags started calling Layton "Taliban Jack."

Afghanistan's Senate has recently called on the government of President Hamid Karzai to open talks with homegrown Taliban militants, in part as a response to civilian casualties caused by combat between the militants and NATO and U.S.-led troops.

Detainee controversy

An explosive political issue in Canada in recent weeks has been the fate of captured Taliban suspects who faced abuse or even torture at the hands of Afghan officials.

When asked how they feel about the treatment of detainees, only 31 per cent say they were outraged and that Canada's reputation has been hurt, while 56 per cent said Canada shouldn't be held responsible for what happens to prisoners held in Afghan-controlled detention centres.

"Basically, Canadians are pretty sanguine about this issue," Donolo said.

The level of outrage was highest in Quebec, at 37 per cent. However, more than half of Bloc Quebecois supporters pronounced themselves outraged, he said.

The public's feelings on the issue stands in contrast with the amount of time spent on it by opposition politicians in Parliament's question period.

Donolo said if the issue was damaging for the minority Conservative government, the damage came more from their perceived handling of it more so than the issue itself.

In the "horse race" portion of the poll released Friday, the Conservatives lost two percentage points in popular support compared to a late April poll (percentage point change from the April 21-24 poll):

  • Conservatives - 34 per cent (-2)
  • Liberals - 31 per cent (+1)
  • NDP - 16 per cent (+3)
  • Bloc Quebecois - 10 per cent (+1)
  • Greens - 9 per cent (-12)

Donolo noted that Canadians hold the military in high regard.

The poll asked about support for public institutions. Eighty-six per cent of respondents say they trust the military, with a net trust rating of +76. Here is how some other institutions compared:

  • Canada Post - 90 per cent, +81
  • RCMP - 80 per cent, +65
  • Canada Border Services Agency - 69 per cent, +53
  • CSIS - 60 per cent, +39
  • House of Commons - 58 per cent, +24

While they may trust the military far more than they do politicians, Canadians are still lukewarm about the Afghan mission.

The poll shows a slight rise in support for the mission compared to a late April poll (percentage point change in brackets):

  • Total support - 40 per cent (+4)
  • Total oppose - 55 per cent (-2)

However, only six per cent say they strongly support the mission, compared to 24 per cent who say they strongly oppose it, Donolo said.

Even with the national rise, those opposed hold at least a bare majority in Quebec, Ontario and the West, he said.

Technical notes

  • This poll was prepared by The Strategic Counsel for CTV and The Globe and Mail.
  • Sampling was carried out between May 14 and 17.
  • One thousand Canadians over the age of 18 were interviewed in a proportionate national sample.
  • The national margin of error is plus or minus 3.1 percentage points.
  • Quebec: 247 sample size, 6.3 percentage-point margin of error
  • Ontario: 379 sample size, 5.0 percentage-point margin of error
  • West: 279 sample size, 5.7 percentage-point margin of error