, 07.22.2019 07:00 AM

The Hidden Conservative Voter

They’re hidden.

They showed up, however, at Brexit. They showed up in the U.S. presidential race in 2016. They showed up in Alberta in 2019, and Ontario in 2018, too.

They’re the THCV – The Hidden Conservative Voter. And they’re changing politics.

June 2016: shocking just about everyone, 52 per cent of Britons voted to leave the European Union. No one really expected that result, including many of those who campaigned for Brexit.

Polls conducted in the years leading up to the Brexit vote consistently showed public opinion split on the EU membership question. A year before the crucial vote, support for the European Union spiked upward, with many more Brits favouring remaining than leaving. That, perhaps, may have been what persuaded then-Prime Minister David Cameron to push for a vote.

It was a critical error, as historians will forever note.

Subsequent vote analysis showed that young Brits favoured remaining in the Union. So did big business, lawyers, economists, scientists and the well-to-do. Voters with lower incomes and fewer higher-education degrees, however, just didn’t.

And they, unlike the young Brits and the others, came out to vote. The “leave” side surged on voting day.

Pollsters and pundits hadn’t seen it coming. Neither did the bookies, even: on the day of the vote, Ladbrokes had been giving six-to-one odds that Brexit would fail.

What happened? Sifting through the Brexit results afterwards, public opinion experts and political scientists saw something they hadn’t previously spotted: what they called, antiseptically, “unrepresentative samples.” In other words, pollsters had too many “stay” voters in their computers – and not nearly enough “leave” voters. That, the British Polling Council determined after a lengthy inquiry, was “the basic problem.”

What is most shocking is that the pollsters repeated their error in the U.S. presidential race, which happened just a few weeks after Brexit. Every single pollster, pretty much, got it wrong. Again.

The New York Times declared Hillary Clinton – who, full disclosure, this writer worked for in three different states in 2016 – had an 85 per cent chance of victory. Huffington Post said she had a 98 per cent chance of winning. The respected poll analyst Nate Silver pegged her chances at 67 per cent – while Princeton University went even further, saying it was 99 per cent.

All wrong, wrong, wrong.

And, as in Brexit, the same thing had happened: pollsters had relied upon unrepresentative samples – allowing Trump voters to hide, in effect. One analyst told GQ that Trump voters hid on purpose: “It may also turn out to be the case that supporters for Donald Trump were shamed into keeping their support quiet. Shy Trump supporters may have kept their support secret from pollsters out of social pressure not to admit their support for a candidate labelled as racist and sexist.”

The same sort of thing has happened in recent Canadian electoral contests. Polls in Alberta suggested the race between Jason Kenney’s United Conservative Party and Rachel Notley’s New Democratic party was far closer than it ended up being. Ditto in Ontario, the year before: mid-campaign polls proclaimed the Andrea Horwath New Democrats had moved ahead of Doug Ford’s Progressive Conservatives. But it wasn’t so: the PC vote surged on voting day, and Ford won a huge majority government.

The moral of the story, here, is clear: pollsters are either missing conservative-leaning voters in their sampling – or those voters are keeping their intentions secret, Until they sit down with a stub of pencil and a ballot, that is.

It’s the THCV – The Hidden Conservative Vote. And it’s changing outcomes in elections across Western democracy.

And for guys like Conservative leader Andrew Scheer, the THCV could be very good news in October.


  1. Davis says:

    > “Every single pollster, pretty much, got it wrong. Again.

    > “The New York Times declared Hillary Clinton – who, full disclosure, this writer worked for in three different states in 2016 – had an 85 per cent chance of victory. Huffington Post said she had a 98 per cent chance of winning. The respected poll analyst Nate Silver pegged her chances at 67 per cent – while Princeton University went even further, saying it was 99 per cent.

    > “All wrong, wrong, wrong.”

    There might be something to your larger point, but this is incorrect. A candidate with a 67 and an 85 or even a 98 per cent chance of winning is not guaranteed to win. This isn’t pedantry. A lot of people have trouble understanding the concept of probability. Of course, if polls repeatedly give underdog status to the eventual winner, their methods and models need to be revised.

    • The Doctor says:

      I totally hear you. Most people don’t understand this statistic and what it actually means. I run across this all the time from alt-right people, who claim that the liberal mainstream media’s official position was that Hillary was a 100% lock to win the 2016 election.

    • Paul says:

      A reasonable interpretation of 98% chance of winning is that if the election was run 50 time, 49 of those times, Hillary would have won. This is highly doubtful and so it is reasonable to say the poll was wrong.

      There are two basic problems with political opinion polling, as distinct from normal random issues, though:
      1. The 95% standard (2 sigma on a normal distribution) is essentially stupid given the number of polls performed.
      2. (and the bigger issue) the pollsters have to create a population model in order to determine how to adjust their non-random sample. Simply put, if you phone 500 people in Ontario at random, you will not get a random sample of political views (technically you could, but it’s overwhelmingly likely that you won’t). So what do you do? Well, you adjust the response rates based on what you calculate as the deviation between your sample and what you think the population is. For example, if you think Ontario is 35% liberal, 25% conservative, 20% NDP, and 20% other, then if you get have the people in your sample saying they are liberals, you will weight their responses at 0.7.
      Since models are often wrong in times of change, such as when you have two additional parties to consider ( the Greens and the PPC), which likely aren’t part of your model ….

  2. J.H. says:

    As in the case of the media, many voters, looking at the political leanings of a lot of the pollsters, prefer to either lie to them or refuse to answer. The lack of trust in both industries has become quite evident the past few years.

  3. J.H. says:

    Here’s a perfect example. https://abacusdata.ca/liberals-and-conservatives-deadlocked-at-32/
    As a voter I know that Abacus is owned in part by Bruce Anderson who is also its chairman. He also is an adviser to the LPC and the father of Kate Purchase, Justin Trudeau’s Dir. of Comms.
    Frank Graves of Ekos was revealed as an adviser and donor to the LPC some time ago.
    CBC was forced to stop using both of the them as pundits and on panels because of those connections.
    Why should I as a voter, trust anything their companies produce in the way of polls?

  4. Jeanbatte says:

    I believe we will see many Conservative voters coming out in the open to help evict the clown from the PMO.

  5. Mark says:

    Except for one thing. Andrew Scheer and the Conservatives are missing the mark on what pollsters (no pun intended given the premise of your column) say is the most important issue to most people: climate change (even a more popular choice than health care, which surprised me). The Cons could well win a plurality of votes and seats, but I’m not so sure about a surge of CPC voters suddenly showing up on voting day.

    • DurhamHCV says:

      Mark, you are falling into the exact trap WK is warning you about – it’s the pollsters and the media who are telling you that Climate Change is the most important issue. THCV will not answer that question or will lie, same as they do about voting intention. I know I certainly do when Forum rings my landline every couple of months.

      • Mark says:

        DHCV, you must have missed where I stated about there being no pun intended… Anyway, I’m aware of your point, but a lot of the folks I hang out with are also telling me that climate change is their primary issue, and none of them except one plan to vote Liberal. They sure as hell won’t vote CPC either. We don’t need pollsters to help us read the mood or opinions of those in our circle of influence.

        • Mark D says:

          I’m a THCV. And yes, I have hidden my voting intentions from pollsters when:

          1) I don’t recognize who they are, or who they are calling me on behalf.

          2) Political campaigns become extreme and divisive and one side resorts to “vote shaming”.

        • Graham Hanlon says:

          Nobody, and I mean nobody I know has climate change anywhere on top of their list of issues, for most it’s an afterthought. Most are concerned with a recession, the affect it will have on the economy, employment and with health care. Honesty & integrity is a concern for some but not all, and finally leadership. Climate change is hardly discussed and then on how much taxes will rise!

          • derew says:

            your right I hope the trudoo nthing is kicked out of our political system no more trudoos they are some sort of evil cult that can bring such hate and division

  6. Douglas W says:

    Hidden Conservative voters: do they have strength in any significant numbers on the Island of Montreal or Toronto’s 416? Suspect they don’t. Which gives the Libs about a 50 seat running head start.Justin’s going to be okay, on election day.

    • Jim Keegan says:

      Those 50 seats are offset by Alberta’s 34 seats and Saskatchewan’s 14. You just lost your running head start.

  7. When people, not necessarily shy CPC supporters, tell pollsters in the mid to high 60s that they want change, you know that Scheer will win, despite himself — or more particularly, his stupid mouth.

    Trudeau and Butts won’t be able to counter that. In fact, because of the general dissatisfaction with this government, expect Liberal election numbers to be comparable to what Dion or Ignatieff got. A fairly good chunk of moderate-to-right LIBERALS who can’t bring themselves to vote for Scheer will stay home, rather than vote for the hapless NDP. Some progressive Liberals will go Green, the flavour of recent months.

    • Douglas W says:

      Scheer needs help from someone else if he’s going to secure the most seats in the October vote. That help may come from an unlikely source: the BQ, who’s been coming on strong in the past six months, polling well outside the Montreal region. If the BQ drains enough support from the Libs in Quebec so that the Libs come out of Quebec with 40 or fewer seats, the incumbent government might be in trouble. Another horse race, like 1972.

      • Douglas,

        I’ve been predicting a Bloc renaissance over at my place but feel somewhat less strongly about that now. My gut tells me, in fact it screams, that the election depends on this:

        May says Greens will work with any party that has a serious plan for the climate.

        Is this genuine or a false flag? As J.H. said earlier, if the Greens are after serious clout in 2019, their only pathway is through the Liberals. What happened the last time May met Trudeau? Was a floor-plan put in? Did they reach a political consensus — or even an informal agreement? I sense Green supporters are about to get royally played here. J.H., quite obviously, wants a CPC majority. I’ve got no double game. I’m in this to send the Liberals the electoral message of a life-time that they will never forget. My problem: by what vehicle do I go about that in Quebec? Hopefully, the polls will reveal the way forward soon enough.

        • That should be at my former place, not my present place.

          • Douglas W says:

            The BQ are going to go hard at Justin. Nasty hard. It’ll mean that the Liberals will fight battles on two fronts: ROC + in Quebec. This is going to grind the Liberals down, and tax whatever ground game and resources they have. As for Green-leaning supporters in 2019, they recall how they put their trust in Justin in 2015, and got played; in their eyes, JT is now shallow, a fake. Greens are going to be game changers. Momentum will be theirs.

          • Ronald O'Dowd says:


            This isn’t Germany or Israel. May’s marker in the sand means either propping up Trudeau or doing something with the NDP and Bloc. How can you be a game changer when your best seat projections, coupled with NDP and BQ numbers don’t come anywhere close to forming even a minority coalition government? That’s why May will inevitably go with the Liberal math. She has no other credible or realistic choice.

  8. Miles Lunn says:

    A more recent example is Australia. The last 50 polls for two years before the election in every single case said Labor Party would win the two party vote, yet when the votes were all counted the L/NP coalition went on to win a larger majority. In Danish and Finnish elections this year, the left won as expected, but the margin was much closer than most polls suggested. Of last provincial elections, UCP and CAQ both clearly outperformed polls; BC Liberals, Saskatchewan Party, Manitoba PCs, and NS PCs got what polls suggested but only in the final days as there was a last minute jump in support in all cases which some pollsters picked up. In Ontario, polls were within margin of error, but PC win was slightly larger than anticipated. In PEI, PCs outperformed polls, but so did Liberals (who were in third), while Greens underperformed (who most polls suggested would win). New Brunswick was bang on although the PCs dropped in the final week with a strong surge for the more right wing People’s Alliance (if this happens federally with Bernier’s People’s Party, Scheer is in trouble).

    So trend does exist, but there are exceptions. This year in Spain, most polls showed the three right wing parties edging out the two left wing but neither having a majority while in the end two left wing edged out the three right wing but no majority. In Newfoundland polls also suggested a PC win, not Liberal, mind you NDP didn’t field candidates in 26 of 40 ridings so I suspect much of that transferred to the Liberals in those ridings thus putting them over the top.

    Where it gets murky is on far right in Europe. When far right is not expected to win, they tend to outperform (see Germany 2017, Norway 2017, Belgium 2019, Finland 2019, Italy 2018, and UK 2015), but when there is a realistic threat of them winning or in case of Spain propping up the government they underperform (see Netherlands 2017, France 2017, Austria 2017, Sweden 2018, and Spain 2019) which suggests to me a lot of the far right votes in Europe are more angry citizens who want to flip the bird to the establishment, but they don’t actually want them to win. With mainstream centre-right parties, it could be partially shy voters as much our news cycle and political establishment are in the large urban centres which do vote overwhelmingly for progressive parties and they are struggling to connect with suburbs and rural areas who don’t share same priorities. Other reason could simply be demographics of who shows up. Generally parties on right do better amongst older voters who are more likely to show up on election day.

  9. The Doctor says:

    Your column somewhat reflects the various theories that are tossed about to explain this. The main ones being:

    1. conservative voters deliberately lie to pollsters

    2. conservative voters “hide from” pollsters

    3. polling outfits are part of the insidious Deep State liberal conspiracy and deliberately make up fake polls to suppress the conservative vote (I’m not kidding, this one is all over Breitbart, for example)

    4. most polling outfits engage in practices that underreport and undersample the conservative vote

    5. related to #4 really: conservative-friendly demographics (e.g., old white people) simply vote more than liberal/left friendly demographics.

    Personally I think 4 and 5 are closer to the truth.

    I notice, for example, that you get markedly different results on a political polling question depending on whether you are polling the population at large, likely voters or registered voters.

    A lot of Trump supporters and alt-righters are very much invested in attacking the polling industry these days, citing 2016 and Brexit as evidence. Problem is, Trump supporters have taken those incidents and blown them up way out of proportion, so now, of course, it’s evidence of yet another Deep State Conspiracy.

  10. Killer Marmot says:

    I suspect that conservatives are, in general, more private, preferring to keep their own counsel. Thus they are less likely to participate in any survey on any matter, introducing significant bias into the results.

  11. whyshouldIsellyourwheat says:

    Liberals only exist in the (financial and housing) bubble economies of the Capital (and regional Capitals) of Panem. Ottawa, Toronto, Montreal, and Vancouver. It is the Hunger Games for everyone else.

  12. Ted Hurley says:

    I am no expert in the numbers but when looking at percentages of support for a specific party, I don’t think it takes into account if they actually vote.

    It is said that Conservatives will walk barefoot across glass in order to vote. The point being that from the polls if the Cons and Libs are statistically tied, the Conservatives will get a higher percentage when the votes come in. More Liberals will say they are voting Liberal but less show up and vote.

    There might be another related effect in our upcoming Fed election. Last election more Libs turned up because they wanted to get rid of Harper. That same feeling won’t be there for Scheer.

    • Ronald O'Dowd says:


      It’s also far from crystal clear how strong the ABT is right now. Will his negatives be as high as Harper’s were?

  13. Derek Pearce says:

    I think one crucial thing you’ve overlooked WK, in both the case of Brexit & Trump, is the “overconfident” small-l liberal: because the polls showed their side with such a comfortable lead / quite high probability of winning, a lot of them didn’t bother to vote, thinking it was not necessary. That is how Brexit squeeked through with such a slim margin and how Trump only won the electoral college but got 3 million fewer votes.
    So, what remains to be seen is how motivates small-l (in addition to large L) liberal voters are to get to the polls whether in advance or on e-day. The polls are fairly close for one thing (or a toss up), and I’m sensing that after seeing Brexit & Trump happen there is a “fool me once shame on you, fool me twice shame on me” feeling among the small-l electorate and that they know better now that they can’t sit on their hands since small-c (and large C) conservatives will overwhelmingly get out and vote.

    • Ronald O'Dowd says:


      The unknown variable is for how many Liberals, like this one, JWR and JP cut ice. That’s precisely why I’m at where I’m at. The more Liberals who sided with them, as I did, the more Trudeau and Butts better start the steady praying.

      • Derek Pearce says:

        I hear ya! Because of his infuriating behaviour in this affair Justin is not getting my vote this year. Perhaps he’ll have done himself in after all– but wow are there ever a tonne of Liberals who hear no evil see no evil speak no evil.

        • Derek,

          The problem with the party system in any democracy is that most everyone “wants” something from the leader or the party/government. They are in it for themselves. That’s why parties are chock full of yes people, boot lickers, ass kissers, sycophants and spineless members — not to mention those same types in cabinet and caucus.

          PET was challenged more than once by his cabinet. Ditto when Day took on Harper in cabinet. And they backed down. But this bunch…history will not be kind to them, to say the least.

    • The Doctor says:

      There’s no question in my mind that some Democratic Party voters stayed home in 2016 because they thought Hillary had it in the bag — because polls seemed to say that she did. Also though there was a huge lack of enthusiasm for her as a candidate, she had the charisma of an armadillo and a number of bitter Bernie supporters also sat on their hands.

  14. David Paterson says:

    I am happy Warren is taking a vacation and happy that he is continuing to post on this blog. Not a good idea to try to do both.

  15. Joseph says:

    Ok its a theory.

    Or it could be that after decades of having politicians promise that they are working for the little guy or the common middle class, but end up with government that makes more laws telling the same folks what they can’t do and what they can’t say, or even laws that compel thought, folks are just ready to vote in a way that will make the governments life difficult.
    All they need is all the indications that a particular candidate or outcome will be the one that will constrict and limit what the government does.
    The indications would be that candidate, party or outcome the experts and pundits attack in a seemingly unified effort.
    Whoever that is, under these circumstances has a good chance of winning.
    Should such an outcome happen, the media and pundits would only confirm it by pursuing a theory that the outcome was rigged, or because someone cheated, or some kind of collusion conspiracy.
    But if you want to go with the hidden voters theory, its likely just as good as mine.

  16. BC Guy says:

    Conservative voters stay quiet as don’t want to hear the screeching from the left who oppose your view and brand you as racist or a homophobe or a nazi etc. It’s ridiculous and is the reason why many conservatives have anonymous twitter handles. The branding by the left is ridiculous

    • Ronald O'Dowd says:

      BC Guy,

      True, but it cuts both ways. Look at AOC: she’s regarded by the right as a traitor, Communist and Antifa. That’s also utter nonsense.

      • BC Guy says:

        I see most of the criticism of her is ridicule of her positions; I have not seen someone calling her antifa and quite frankly her economic positions are very far left that she opens herself up to that criticism. Conservatives are branded as racists just because….with no basis at all

        Cripes …the Trudeau govt has done that too if people have a concern about illegal immigrants…It’s despicable

        • Fred from BC says:

          Agreed. Conservatives don’t run around vandalizing your property if you put up an election sign for the NDP. They don’t don black masks, arm themselves with bricks and clubs and show up at a Naomi Klein speaking engagement to physically attack anyone who dares hear her speak. They don’t stage 350+ incidents of fake attacks on white people….etc, etc.

          And the worst part? They don’t appear to be learning from those mistakes. They are going to be instrumental in re-electing Donald Trump in the US and electing Andrew Scheer in Canada, followed by much wailing and gnashing of teeth as they flail about blaming racists, bigots, ‘homophobes’, misogynists, white supremacists and all their other favorite scapegoats. They just can’t help themselves, can they?

          • BC Guy says:

            Great post. The whole thing is so tiring and juvenile. Their constant rants about racists and nazis are sickening and divisive. Butts calling people who ridiculed Trudeau goofy peoplekind “alt right nazis” a perfect example

        • BC Guy,

          I agree with you on immigration. Having legitimate concerns is not the same ballgame at all as wanting to deliberately exclude, often with no or the most frivolous of motives. Some might even say Trumped-Up Motives.

          • Fred from BC says:

            “I agree with you on immigration. Having legitimate concerns is not the same ballgame at all as wanting to deliberately exclude, ”

            Yes, and the pollsters are now verifying that even those who identify as Liberal or NDP supporters are expressing their concerns about illegal immigration. In the past, it was only Conservatives who dared express that opinion, wasn’t it?

  17. Pedant says:

    The New York Times declared Hillary Clinton – who, full disclosure, this writer worked for in three different states in 2016 – had an 85 per cent chance of victory. Huffington Post said she had a 98 per cent chance of winning. The respected poll analyst Nate Silver pegged her chances at 67 per cent – while Princeton University went even further, saying it was 99 per cent.

    All wrong, wrong, wrong.


    Probability is neither wrong nor right. I especially wouldn’t classify Nate Silver’s 67% as “wrong”. Up until the last moment, it was perfectly reasonable to say that Clinton had a 2/3 chance of winning. Everything had to go just right in several states for the other 1/3 to effectuate, and that’s exactly what happened. I will note that the popular vote pre-election polls were mostly correct.

    • Vancouverois says:

      I will note that the popular vote pre-election polls were mostly correct.

      Well, isn’t that the main point?

      These probabilities were all based on polling that turned out to be very much less than accurate. They were wrong; they were BADLY wrong.

      I don’t think you can’ brush that aside simply by reiterating that they were never claiming to be one hundred percent certain.

  18. Brendan Kane says:

    “the latter absolutely in AB and with Beto vs Cruz”

    18 of the last 19 polls had Cruz winning by 3% or more. He won by 2.56%

Leave a Reply

Your email address will not be published. Required fields are marked *