Canada-United States Election Conversions

[Thande]

I thought I would create a separate thread for this, an idea that came to me yesterday.

People have tried to do wikiboxes before that are Canada as the US or vice versa, of course, but I thought it might be interesting to create a consistent methodology into which one can plug numbers that will spit out hard results. Obviously there's still a lot of handwavium involved, but I thought it was a project worth trying.

The most obvious place to start is to find a way to make numbers equate between the two countries. I don't mean the absolute population (in which the US obviously has nearly ten times Canada's population) but the relative population between units. Canada currently has 13 provinces and territories, although for reasons that should become clear, I am working with the situation in the early 90s, before Nunavut had split off, so we will call it 12 (and we can always just recombine the numbers for the post-1995 Northwest Territories and Nunavut back into one for later elections). The US has fifty states and one district which has a vote in presidential elections. How to make these equate?

I could just try different combinations of states that would equate as closely as possible to the relative population share of a province. For example, Quebec in the 1991 census (the one I am using) had 25.3% of Canada's population, so I could just pick the US states that add up to the closest figure I can get for 25.3% of the USA's population. However, that wouldn't be a very good analogy if it was just random states strewn all over the map. These 'provincial analogies' need to be as analogous as possible, which at a bare minimum means they need to be a coherent lump of states. Ideally, that 'lump' should also be culturally comparable between the corresponding province, or rather, it should play a comparable role in US politics as the Canadian prototype does in Canadian politics. When this analogy is sufficiently strong, I have allowed this to introduce more deviation between the strict percentages. This is most obviously apparent in two cases from the decisions I have made: the 'Left Coast' makes up a bigger proportion of the USA than British Columbia does of Canada (California alone is almost identical to BC's share, but WA and OR really don't fit anywhere else); and New England is the obvious analogue to the Maritimes, but whereas the Maritimes make up 8.6% of Canada (or did in 1991), New England made up only 5.3% of the US. So the Maritime provinces are a tad underrepresented, unfortunately, as I couldn't exactly hack off half of upstate New York and add that in. I did consider making Newfoundland Hawaii instead due to the 'joined later' aspect, but that seemed a bit of a stretch.

So, here are the analogues I have chosen and my reasoning behind them.



Comparison of share of population of these state combinations in the US (left, 1990 census) and share of population of the original provinces in Canada (right, 1991 census):


My reasoning:

Alberta: I originally looked at making this a slice of the Plains instead, but Alberta makes up such a relatively huge share of Canada that it got ridiculous, stretching from Montana to Oklahoma and beyond. Instead, well, Texas and Alberta are often compared (conservative, cattle, oil, etc.) so I based it on that and added in Kansas for the agrarian side of things. There is no really direct analogue for the Hispanic aspect to the US Southwest in Canada so I have mostly just ignored that, unless you want to make tenuous Métis comparisons.

British Columbia: Obviously culturally analogous to the 'Left Coast', hence the whole 'Cascadia' idea--as noted above, this does mean it got slightly overrepresented in the Canadian-US.

Manitoba: Agrarian Plains farming, mining, and lots of nothing analogous to the northern part of the province. Squeezing Colorado and Utah in are a bit iffy but I don't know where else they'd go.

Newfoundland & Labrador: Maritime. A bit underrepresented, but not much else one could do. Maine's split and Vermont's independent republic are the closest I could get to NL's late addition to Canada.

New Brunswick: Again, a bit underrepresented, but a loose fit.

Northwest Territories: Not really similar geographically, but the significant native element means there is a line that can be drawn.

Nova Scotia: The most significant part of the Maritimes just as MA is the most significant part of New England. Halifax equals Boston here.

Ontario: Overly large and complex, dominant, with the big city everyone else in the province loves to hate (Toronto/NYC). Industrial, the traditional centre of power. Obviously the fact that these states wrap around Ontario itself via the Great Lakes means there is some cultural comparison. There isn't really a huge hinterland like Ontario's north, but the fact that population is tilted towards the east coast loosely resembles the concentration in southern Ontario.

Prince Edward Island: The small island everyone forgets = the other small island everyone forgets.

Quebec: The South is often the analogy for Quebec chosen in these setups, and one can see why, despite the obvious dissimilarities as well. A history of separatism (er, yes), does its own thing, often ends up being politically dominant on a national level despite being in the minority, gets its own way despite theoretically being on the losing side of history (Wolfe vs Grant). It's not quite the same, in that the South exercised its power mostly through Congress and didn't supply as many Presidents as Quebec has Prime Ministers, but you get the idea. This is also a slightly iffy definition of the South to make the numbers work, and it does have the disadvantage that it implies a connection with Texa/Alberta that doesn't exist in the Canadian prototype. In a broader sense, Ontario-Quebec kind of works as an analogy to how the US became independent as a north and south that had an obvious cultural divide from the start but worked together against a common foe, the former's culture was more successful at expanding into the interior than the latter's and the latter ended up feeling like an embattled minority, the latter is more agrarian and has more of a tendency to romanticise a traditional lifestyle, etc. Note also how the capital also straddles the line between them!

Saskatchewan: More agrarian Plains stuff, combined with a political culture that is mostly conservative nowadays but historically flirted with radicalism. (Of course, there was a direct exchange of ideas between Saskatchewan and these states in the past).

Yukon: Another pretty obvious one.

 

[Thande]

Right, what can we do with these analogies once they're made? Well, one simple thing is to show a US election in Canada, because all you have to do is add up the votes for the relevent states. Because I used the 1990/1991 census...es, why not the unusual US election of 1992? Orange is Perot here.



OK, that's kind of interesting, and slightly resembles a Liberal landslide in the mid-20th century if you reverse the colours. Quebec sticks out a bit due to the US South realigning towards the Republicans at this point, but it kind of works if you imagine the Socreds folding into the Tories and the Bloc not forming.

It doesn't tell you that much, though. What would be more interesting is to see a Canadian election on the US. Well, I could just do the same in reverse, take the Canadian numbers for each province and fit them to the analogous US 'provinces' I've invented. But that feels like a cop-out, because we don't want to see a US election result on arbitrary divisions we've just made up, we want to see them on the actual states. So this will require more thought.

 

[Uhura's Mazda]

Interesting stuff. For bonus points, bring in a three-way analogy with Australian states.

For all that we take pains to acknowledge that Canada is very different fro America, it's fascinating to see the contiguous communities of 16th-17th century British planned colonies in NE/Maritimes, bougie wankery in Cascadia, and of course Mumbypunk in Sask/Midwest.

 

[Lord Roem]

Interesting stuff. For bonus points, bring in a three-way analogy with Australian states.

For all that we take pains to acknowledge that Canada is very different fro America, it's fascinating to see the contiguous communities of 16th-17th century British planned colonies in NE/Maritimes, bougie wankery in Cascadia, and of course Mumbypunk in Sask/Midwest.

The exiled Old Believers as Canadian Mormons would be interesting to see - Yukon as *Deseret but it's all Two Fingered Signs of the Cross.

 

[Dan1988]

That's absolutely brilliant thinking, Thande. It makes a lot more sense than the other divisions of the US I've seen, even with the stretches you've made to make it all fit (and as for the Rhode Island = PEI comparison - I never thought of it that way before, but it does make sense once you get out of the cities).

 

[Thande]

This was way more complicated than even I thought, so I will do it a bit at a time.

What I have done is compare the 1993 Canadian election to the 1992 US election. This works better than your average US election because of Perot's third force. If you compare the numbers, a few things do leap out, especially if you count Reform together with the Bloc (which works because Reform didn't stand in QC and the Bloc didn't stand outside it)



(that says 'Reform/Bloc' cut off).

Liberal almost = Democratic, PC almost = Perot, and Reform+NDP roughly = Republican. This is more going for close numbers we can crunch rather than implying any ideological parallel, although it is aesthetically pleasing to put the Republicans in the same box as Reform and effectively to imply that in the South they are appealing to old secessionist impulses (see, it kinda works).

I can put these together and come up with corrective factors--not for the whole country but for each 'province' made up of bashed-together states--and then try to model the election result in each state. The idea is, for example, if the area of "Alberta" was 36.7% Democratic in the OTL US election, but the real Alberta was 25.1% Liberal, then for each state we can apply a corrective factor of -11.7% to its OTL Democratic figure to produce a unique Liberal number for each one. Doing Reform and NDP is slightly trickier because you have to split them up first using the balance between the two parties for each province. It's a bit murky, but it's something to go on.

So let's indeed start with the states making up the province of "Alberta":

Kansas
Reform 50.8%
Liberal 22.1%
PC 19.2%
NDP 4.2%
Other 4.0%

Now let's stop for a moment - does that pass the acid test of what somewhere like Kansas would 'look' like, to our gut feeling, in a Canadian political context? I'd argue yes--it nicely fits things like the Orman case a few years ago where there's an obvious split in the opposition, that no matter how dominant the right is, much of the opposition takes the form of the centre-right as well, etc.

Moving on:

New Mexico
Reform 49.4%
Liberal 34.2%
PC 8.3%
NDP 4.1%
Other 4.3%

What about this one? Not quite as convincing, perhaps, but you can imagine a big anti-immigration vote fuelling Reform here, while the Liberals get the Hispanic vote, and nobody much cares about the PCs and Dippers.

Oklahoma
Reform 54.3%
Liberal 22.3%
PC 15.2%
NDP 4.0%
Other 3.9%

Bit hard to answer this one given the sheer dearth of examples of third parties in Oklahoman history...maybe the Liberal vote here is associated with Native Americans?

Texas
Reform 52.4%
Liberal 25.4%
PC 14.2%
NDP 4.4%
Other 4.0%

Yeah, sounds about right.

If you're thinking these figures seem a bit similar--that's in part because the states weren't that far apart with the original figures, either. It might be better to introduce a scaling factor to take into account different populations, come to think of it, but I'm not sure how that would be possible.

 

[Thande]

British Columbia, one of the NDP's stronger regions:

California
Reform 36.4%
Liberal 28.9%
NDP 15.5%
PC 12.6%
Other 6.6%

What, I hear you cry, Preston Manning winning the Golden State? Well, the early 90s were a different world--remember, 1992 in OTL was the first election since the 1964 landslide that California had voted Democratic. And of course it's a very split vote here.

Oregon
Reform 36.4%
Liberal 25.3%
PC 16.2%
NDP 15.5%
Other 6.6%

I checked the spreadsheet formula for this one three times because the NDP and Reform figures are the same as California, but it's just thatthe Republican voteshare in the two states was almost identical in OTL.

Washington
Reform 36.0%
Liberal 26.3%
PC 15.7%
NDP 15.3%
Other 6.8%

Remember even by Canadian standards the 1993 election was very atypical--I may use the same methodology to look at other elections in future if I can come up with a suitable formula for working it out from a more conventional two-party US election.

 

[Thande]

Manitoba now.

Arizona
Liberal 46.2%
Reform 22.5%
NDP 16.8%
PC 10.8%
Other 3.7%

Doesn't feel quite right that New Mexico goes Reform while Arizona goes Liberal, but it is the sort of thing that routinely goes on in Canada between say Manitoba and Saskatchewan.

Colorado
Liberal 49.8%
Reform 21.1%
NDP 15.7%
PC 10.3%
Other 3.1%

I'd like to see where the NDP gets in a more normal (by Canadian standards) election.

Idaho
Liberal 38.1%
Reform 24.6%
NDP 18.3%
PC 14.1%
Other 4.9%

These results look unrealistic to us now, but across the whole area of my "Manitoba" in the actual US presidential election of '92, the Republicans were only ahead of the Democrats by by 2.8%, and because I've split their vote between Reform and the NDP...

Montana
Liberal 47.3%
Reform 20.6%
NDP 15.4%
PC 13.1%
Other 3.6%

Fun fact, the only time I went to Montana was in mid-1994, at which point it still had a Democratic Representative...

Nevada
Liberal 47.0%
Reform 20.4%
NDP 15.2%
PC 13.2%
Other 4.2%

Nevada has changed a lot politically since this time - Clinton only barely carried it in a three-cornered race.

Utah
Liberal 34.3%
Reform 25.3%
NDP 18.9%
PC 14.4%
Other 7.1%

Utah going Liberal? But note that Reform+PC would be a conservative majority. 1993 was a weird election.

Wyoming
Liberal 43.6%
Reform 23.2%
NDP 17.3%
PC 12.6%
Other 3.4%

Another pattern that seems strange now, but again, Bush carried Wyoming by less than 5% in 1992.

 

[Thande]

Newfoundland & Labrador is weird because both the NDP and Reform had almost zero presence there in OTL 1993. The corresponding US "province" I've carved out was also one of Perot's best areas, so weirdly Perot's score was almost the same as the PCs' score even without any correction (26.2% vs 26.7%). Effectively what I've done is almost just add most of Bush's votes to Clinton's, as the Liberals were absurdly dominant in NL in 1993.

Maine
Liberal 65.9%
PC 31.0%
NDP 1.5%
Reform 0.5%
Other 2.3%

I wonder if I can do 96/96, that might be an interesting contrast.

New Hampshire
Liberal 66.0%
PC 23.1%
NDP 7.1%
Reform 2.1%
Other 1.6%

New Hampshire's bigger Republican vote than the other two states' results in it being the best area for both Reform and the NDP, which does 'feel' right if you think about what New Hampshire is like.

Vermont
Liberal 73.2%
PC 23.3%
NDP 1.5%
Reform 0.5%
Other 3.1%

I've wondered if Vermont would be a good place for the NDP in a modern analogy, but ironically in this setup it's as dominated by the Liberals in 1993 as it is by the Democrats in 2018 today (whereas it wasn't dominated by the Democrats in 1992--while it went fairly strongly Democratic that year, it was also the first time it had gone Democratic in decades).

That's enough of that for now but I may continue it in future.

 

[OwenM]

Very interesting idea.

 

[Thande]

Progress so far, using a new colour scheme what I have just made up.

 

[Thande]

Of course muggins here forgot the next three provinces/territories (alphabetically) are all dead simple because they're only one state each, so why on earth did I decide to stop before them...in other words, these are just the OTL provincial results, so no number-crunching required.

New Brunswick
Connecticut
Liberal 56.0%
PC 27.9%
Reform 8.5%
NDP 4.9%
Other 2.7%

Northwest Territories
Hawaii
Liberal 65.4%
PC 16.2%
Reform 8.4%
NDP 7.7%
Other 2.3%

Nova Scotia
Massachusetts
Liberal 52.0%
PC 23.5%
Reform 13.3%
NDP 6.8%
Other 4.4%

 

[Thande]

So Ontario then.

Delaware
Liberal 51.2%
PC 20.3%
Reform 19.3%
NDP 5.7%
Other 3.5%

Sounds about right

District of Columbia
Liberal 92.4%
PC 4.1%
NDP MELON MELON MELON OUT OF CHEESE ERROR REDO FROM START

No really, DC actually breaks maths because of how absurdly dominant the Democrats are - the NDP and Reform Party end up with negative scores so Other can have 4.8%. Let's just call it a 50% majority shade, carefully back away, and move on.

Illinois
Liberal 56.3%
Reform 18.5%
PC 16.5%
NDP 5.5%
Other 3.2%

Instinctively Illinois does feel like the sort of place where the right would split and let the centre-left machine get even more dominant.

Indiana
Liberal 44.5%
Reform 25.1%
PC 19.6%
NDP 7.5%
Other 3.3%

Much closer, and a united conservative party would beat the Liberals, albeit only just.

Maryland
Liberal 57.5%
Reform 19.5%
PC 14.0%
NDP 5.8%
Other 3.2%

Of course, this comes from a time when the Democrats were less dominant in Maryland anyway.

Michigan
Liberal 51.5%
Reform 20.1%
PC 19.1%
NDP 6.0%
Other 3.3%

Detroit is Windsor in this scenario I suppose, although then the NDP would be higher...well, if it was today.

New Jersey
Liberal 50.7%
Reform 23.3%
PC 15.4%
NDP 6.9%
Other 3.6%

It was trending towards the Democrats at the time, I recall.

New York
Liberal 57.5%
Reform 18.2%
PC 15.6%
NDP 5.4%
Other 3.4%

The Liberal Party of New York would have its job cut out to nullify the result of provincial elections with those numbers, one assumes, but I'm sure they'd try anyway.

Ohio
Liberal 47.9%
Reform 21.6%
PC 20.8%
NDP 6.4%
Other 3.3%

Would be a swing state if the right was united.

Pennsylvania
Liberal 52.9%
Reform 19.9%
PC 18.0%
NDP 5.9%
Other 3.3%

Similar to Ohio but more tilted towards the Liberals.

West Virginia
Liberal 56.1%
Reform 19.3%
PC 15.7%
NDP 5.7%
Other 3.1%

WV was one of the more Democratic states in this group in 1992, prepare for narrator voice.

Wisconsin
Liberal 48.9%
PC 21.3%
Reform 20.4%
NDP 6.1%
Other 3.4%

Jean Charest would have campaigned in etc. etc.

 

[Roger II]

District of Columbia
Liberal 92.4%
PC 4.1%
NDP MELON MELON MELON OUT OF CHEESE ERROR REDO FROM START

I still think a Green opposition party would do well in DC if it campaigned on the right issues*. If you want slightly less maths-breaking results, maybe lump independents with a party somewhere? I know it's semi-common to run as an independent and do well.

*Either being pro-environment and pro-transit and reduced driving, as I would like, or shameless giveaways to NIMBYs, as would do well.

 

[Thande]

Prince Edward Island is easy because again there's only one so it's just the OTL result
Rhode Island
Liberal 60.1%
PC 32.0%
NDP 5.2%
Reform 1.0%
Other 1.7%

Now, Quebec.

Alabama
BQ 54.6%
Liberal 31.3%
PC 9.7%
NDP 1.7%
Other 2.7%

Reminds me a bit of those "Jeb! Bush runs as a third party in 2016" things, with him as the PCs...

Arkansas
Liberal 43.6%
BQ 42.8%
PC 9.3%
NDP 1.3%
Other 3.0%

Clinton/Chrétien's 'home state effect' (see how this works?) makes a difference here.

Florida
BQ 48.1%
Liberal 29.4%
PC 18.7%
NDP 1.5%
Other 2.4%

This does kind of feel right to my mind - the idea that the Southern Bloc is not as dominant here as elsewhere due to the snowbirds and so on, but they end up voting PC instead of Liberal so it's a split opposition.

Georgia
BQ 50.0%
Liberal 33.9%
PC 12.2%
NDP 1.6%
Other 2.4%

Again, feels a bit Wallace '68.

Kentucky
BQ 48.5%
Liberal 34.9%
PC 12.5%
NDP 1.5%
Other 2.5%

I sort of instinctively feel this should stick out more, but then the Upper South didn't stick out as much at this point I think.

Louisiana
BQ 48.1%
Liberal 36.0%
PC 10.7%
NDP 1.5%
Other 3.7%

Ironically the only place that might vote for the actual BQ is one where the Liberals are relatively close here.

Mississippi
BQ 56.6%
Liberal 31.1%
PC 7.6%
NDP 1.8%
Other 2.9%

The strongest vote for the Southern Bloc here, which again feels right.

Missouri
BQ 41.3%
Liberal 34.4%
PC 20.6%
NDP 1.3%
Other 2.4%

Despite its historical role, Missouri doesn't really belong in "The South" and this is sort of visible here--the smallest vote for the Bloc, even though they still top the polls, and the much more significant vote for the PCs.

North Carolina
BQ 50.5%
Liberal 33.0%
PC 12.6%
NDP 1.6%
Other 2.3%

I feel as though NC should have a weaker Bloc and a stronger PC and Liberal vote than SC, let's see...

South Carolina
BQ 55.0%
Liberal 30.3%
PC 10.4%
NDP 1.7%
Other 2.6%

Yep.

Tennessee
BQ 49.6%
Liberal 37.5%
PC 8.9%
NDP 1.5%
Other 2.5%

If this was truly analogous, I think the PCs would be higher in Tennessee--East Tennessee would probably be sticking with them, whereas the Republican vote has got turned into the Bloc vote here.

Virginia
BQ 52.0%
Liberal 31.0%
PC 12.5%
NDP 1.6%
Other 2.9%

Doesn't look right nowadays, but again, Virginia's changed a lot since '92.

 

[Thande]

Final stretch

Saskatchewan

Iowa
Liberal 35.1%
Reform 27.5%
NDP 26.9%
PC 7.7%
Other 3.4%

A tight three-way race! Reflecting the state of farming at the time in both the US and Canada.

Minnesota
Liberal 35.3%
Reform 24.7%
NDP 24.2%
PC 13.0%
Other 2.8%

NDP would probably be higher if this was more realistic.

Nebraska
Reform 32.2%
NDP 31.4%
Liberal 21.2%
PC 12.7%
Other 3.8%

This on the other hand feels quite convincing.

North Dakota
Reform 31.0%
NDP 30.3%
Liberal 24.0%
PC 12.1%
Other 3.3%

Of course, the NDP's roots are in North Dakota...

South Dakota
Reform 29.2%
Liberal 28.9%
NDP 28.%
PC 10.8%
Other 4.2%

I take back what I said above - that's not a tight three-way race, THIS is a tight three-way race! Wow.



And finally, Yukon = Alaska

Yukon
NDP 43.3%
Liberal 23.3%
PC 17.8%
Reform 13.1%
Other 2.5%


And done.

 

[Thande]

And here's the map.



I kind of want to try other elections now, but not sure how to do them. Maybe 96/97, but that would need a bigger corrective factor if 8% of Perot is doing duty for 18% of PC.

 

[Dan1988]

Prince Edward Island is easy because again there's only one so it's just the OTL result
Rhode Island
Liberal 60.1%
PC 32.0%
NDP 5.2%
Reform 1.0%
Other 1.7%

This result I can imagine - IOTL this was around the time the Mafia was getting busted (Raymond Patriarca was arrested a couple of years before that by the FBI, so around here would be the trial phase), plus Bruce Sundlun being, well, Bruce Sundlun. The Reform percentage here also works since Gov. DiPrete (which I'm assuming here would be RI's first Reform Governor after a string of alternation between Liberal and PC Governors) was involved in major financial scandals early on and this would be appropriate punishment. Among other things pertaining to local factors. As for the placement of the PC Party within an Atlantic Canada/Maritime context here as elsewhere in New England makes sense since the local GOP (well, most of the time) and the PC Party in its Canadian equivalent are ideologically similar. (Yes, I know, that bit disregards how you're using the GOP for numbers crunching - just chiming in based on what I know.)

 

[Bruno]

Jeez - that's fantastic work, Thande.

 

[Upvoteanthology]

This is great! The way you’ve divided the US based on Canadian population proportions reminds me of this map I made just a couple months ago: