Can Ontario's election tell us anything about the federal race in 2019?

It's an old theory about Ontario politics that voters tend to electprovincial governments of a political hue different from the one running the show in Ottawa.

It's also one of those rare instances in politics of the conventional wisdom being (mostly) right.

It is impossible to say whether this pattern one party in power federally, another party in power in Queen's Park is by design or merely coincidence. But it isn't a two-way street. While the pattern bodes illfor Kathleen Wynne's Liberals in the June 7 provincial election, you can't really draw any conclusions from it aboutJustin Trudeau's re-election chances in 2019.

• Ontario Poll Tracker: Follow the latest poll numbers
• Ontario Votes 2018: Full coverage of the election campaign

If the only thing we knew about the Ontario election was the name of the party governing the country, betting against Wynne's Liberals would make sense. In 78 per cent of provincial elections held in Ontario since 1867, the winning party was ideologically different from the party whose leader was sitting in the prime minister's office in Ottawa at the time.

That figure increases to over 90 per cent for elections held since 1943.

The most recent exception was in 2003, when the Ontario Liberals ousted the Progressive Conservatives from power while Jean Chrtien was still the Liberal prime minister. But that camejust a year before the federal Liberals themselves were reduced to a minority government.

The only other exception in the last 75 years was in 1959, when the Ontario PCs won at a time when John Diefenbaker's PCs were in power in Ottawa. Diefenbaker'sstint inoffice proved relatively short, while at thetime the Ontario PCs were in the 16th year of what would be a 42-year span in government.

It's hard to tell whether Ontario voters are actively choosing to establish checks and balances between Ottawa and Toronto. It could be just a coincidental product of electoral cycles, withthe Liberals in Ottawa outlastingtheir welcome just as the PCs outlasttheirs in Toronto (give or take a few years), and vice versa.

But the omens aren't good for Wynne, even if you put asideher party's deepdeficit in the polls.

Federal elections show no Toronto-Ottawa split

But thehistory alone shouldn't lead Trudeau's Liberals to believe aWynne defeat improvestheir oddsin next year's federal election.

There is no apparent relationship between the party governing Ontario and the electoral performance of its federal cousin. In 52 per cent of elections, the federal party that won the most seats in Ontario wasdifferent from the one governing the province at the time. In the other 48 per cent of cases, the provincial cousin of thefederal party that won Ontario was also in power in Queen's Park.

A Toronto-Ottawa split in federal elections has been more common lately; two-thirds of federal elections since 1962 saw one party take the most federal seats in Ontario while its political rival was in power in Toronto. But that pattern in federal elections is still significantly weaker than it is in provincial elections over that period.

Still, the results of the Ontario election will be watched closely by federal party strategists for lessons about what works and what doesn't with voters. They can also take a look at riding-level results to get a clue about where they might have their best chances of a breakthrough in 2019.

A tale of two Liberal parties

With the exception of northern Ontario, provincial and federal ridings in the province follow the same boundaries. That makes it possible to directly compare election results. Do that for the last three electionsand you do see a riding-level relationship between how a federalparty doescompared with its provincial counterpart though it isn't a perfect one.

Of the three major parties, therelationship between the electoral performances of the federal and Ontario Liberal parties appears to be the weakest. That's likely a result of a series of roller coaster elections for the provincial Liberals a big provincial majorityin 2007, a minority in 2011, another majority in2014 as the federal party suffered a historic defeat in 2011 and an equally historic surge in 2015.

The difference between the performance of the Ontario Liberals in 2014 and the federal Liberals in 2015 averaged +/- 6.6 points per riding;it was +/- 12.7 points per riding in 2011.

Nevertheless, the better a Liberal candidate did at the provincial level, the better their Liberalcounterpart was likely to do at the federal level in that riding compared to other parts of the province a pattern that applies to the other parties as well.

That would seem to go against the history of Ontario voters punishing a party at one level of government but rewarding it at the other. But the riding-level data showhow a party's strongest and weakest regions of the province remain relatively constant, win or lose.

Each party's voters can be generally found in the same corners of Ontario from election to election, whether it's federal or provincial. The only thing that changes is the number of votes the party pulls in.

PCs, Conservatives show closer relationship

The relationship between the performances of the Ontario PCs and the federal Conservatives in recent elections is much stronger than it is for the Liberals, averaging as little as +/- 5.9 points per riding in 2015.

This could be due to voters in Ontario being more apt to switch between the Liberals and New Democratsthan they are to swing between the PCs and the two other parties. Polls routinely show that PC voters are the least likely to have a second choice.

NDP pattern shows importance of local candidates

Though the relationship between the federal and provincial New Democrats the only one of the three parties that has official ties between its two wings weakened in 2014-15, it was much stronger in 2007-08 and 2011, with an average difference of +/- 3.9 and4.2 points per riding.

That relationship is particularly close in ridings where the New Democrats haven't had much support, while it's weaker where the party has popular local candidates. This is generally the story for parties that have less of a shot of forming government; they need strong MPsand MPPs to withstand voters' tendency to want to back a winner.

That may be less of a factor this time as the New Democrats move into second place in the polls.

The Ontario NDP hasoutperformed the federal party in parts of southwestern Ontario where the provincial party made gainsin 2014, while the federal New Democrats did better in ridings like Ottawa Centre and (until 2014) Windsor and parts of southwestern Ontario that foreshadowed subsequent provincial breakthroughs.

Swingingrelationship

But while the results of one election in Ontario may show where a party can make gains in a federal election in the province, it doesn't say much about how well a party will do overall. In the last three federal elections, provincial election performance would have predicted a federal party's score to within three points in just 29per cent of ridings.

If the difference in support between the provincial and federal parties provincewide is applied to those ridings, however, it is possible to then predict the federal party's result to within three points in 42 per cent of them an indication of how a party's regional strength exists at both levels of government but is still relative to how it's doing overall.

So there will be many lessons to learn from the next few weeks of campaigning in Ontario, but there's a limit to what one election can say about the next even in a province where an old theory still has legs.