The Conservative Party has lost more support, a new poll has found -- and that can be tied to a popularity drop for its leader Stephen Harper.

Nationally, the numbers break out like this (the May 8 figure is in brackets):

• Liberals: 34 per cent (27), +7
• Conservatives: 26 per cent (31), -5
• NDP: 19 per cent (20), -1
• Greens: Nine per cent (7), +2
• Bloc Quebecois: 13 per cent (14), -1

Stephen Harper was viewed favourably by 50 per cent of respondents on May 8. In the June 9 poll, that dropped to 40 per cent -- a 10-point decline.

"The deal with the Bloc to bring down the government wasn't very popular," Tim Woolstencroft -- managing partner for The Strategic Counsel, which conducted the poll for CTV and The Globe and Mail -- told CTV.ca.

The Gurmant Grewal affair -- in which the Conservative MP claimed to try and "sting" the Liberals into buying his and his MP wife's support for the May 19 budget votes -- likely hurt the Tories, he said.

"But frankly, as Canadians get to know Harper, they (think) he might be a little lacking in leadership," he said.

In the Prairies, a Conservative bastion, Harper went from a favourability rating of 63 per cent down to 44 per cent -- a 19-point drop. But he lost approval in every region.

Among Conservative voters in general, he went from 88 to 76 per cent approval -- a 12-point drop. With "hard" Conservatives, he dropped from 92 to 79 per cent -- a 13-point drop.

"I think this represents disappointment with his failure to bring down the government," Woolstencroft said.

In comparison, Prime Minister Paul Martin went up two points, from 42 to 44 per cent.

NDP Leader Jack Layton and Bloc Quebecois Leader Gilles Duceppe both dropped two points, but remain very popular. Layton is seen favourably by 61 per cent of voters while Duceppe is liked by 70 per cent

Issues

The sponsorship scandal has been replaced by concerns about government stability as the top issue facing the country.

Sponsorship went from 23 per cent to 11, a drop of 12 points.

Government leadership and stability was the top issue for only eight per cent of respondents on May 8. That figure rose to 18 per cent on June 9 -- a 10-point jump, or more than double.

Medicare was the top concern of 15 per cent, although the June 9 Supreme Court ruling on the constitutionality of private health care in Quebec had just come out.

The rise of leadership as a concern "is reflecting a general concern about the quality of political leadership in the country," Woolstencroft said.

While people might not like the Liberals' tactics for holding onto power, "they don't see any alternative," he said.

Holding pattern

In an Ipsos-Reid poll released June 9, 2004, during the heat of the federal election campaign, the Liberals led the Tories 32 per cent to 31 per cent.

When the vote was held on June 28, the Liberals captured almost 37 per cent of the vote, while the Conservatives had almost 30 per cent.

Despite all the political turmoil in the interim, the 34-27 margin enjoyed by the Liberals is slightly better than it was in the Ipsos-Reid poll of a year ago.

"I think the only thing that would change these numbers significantly is a change in leadership, either by the Conservatives or the Liberals," Woolstencroft said.

"Clearly, these numbers are largely shaped by the key leadership players, and the public has serious misgivings about both of them," he said.

While a new issue might shake the numbers up, it likely won't have an impact for more than three or four weeks, he said.

The public has probably absorbed the impact of the sponsorship scandal, and Woolstencroft said he didn't expect the release of Justice John Gomery's reports will have much impact on public opinion when released late this fall.

Another issue is the Conservatives must learn how to make themselves more appealing to urban voters. Woolstencroft said the Conservatives have only nine per cent support in Canada's three largest urban areas.

"You can't be competitive and think you can form a government when you're at nine per cent ... trailing the Liberals and NDP by significant margins," he said.

Technical notes

The findings are based on interviews conducted by telephone among a national sample of 1,000 adult Canadians 18 years of age or older. The interviews were conducted from June 6 to 9, 2005. A sample of 1,000 yields a margin of error of +/- 3.1 percentage points, 19 times out of 20.