Warning: This is a rush transcript generated automatically and may contain errors.
Asbed: Hello everyone and welcome to this Conversations on Groong episode. We have with us Hrant Mikaelian, a political scientist and multidisciplinary researcher in social sciences based in Yerevan. We are going to discuss the notoriously bad political polls in Armenia as well as the developing aftermath of the parliamentary elections 11 days ago on June 7. Hello Hrant, always good to have you on the Groong podcast.
Hrant: Hello Asbed, thank you for having me.
Hovik: Thank you for joining us and just in terms of point of order we’re going to talk about the election aftermath and then we will discuss polling in the end I want to begin by asking you know by getting your thoughts on what happened on June 7th we we know now that by now the Central Electoral Commission, managed by one of Pashinyan’s allies, if that’s even an understatement, rubber-stamped the results of a falsified election were, in my opinion, falsified, but I think in the opposition’s opinion as well, where Prosperous Armenia was robbed of its votes.
What we’re seeing is essentially the only path left for the opposition is to go to the Constitutional Court. and we know the makeup of that court already. To add fuel to the fire, Pashinyan has not actually set a new conciliatory tone of trying to have solidarity. In fact he has increased his rhetoric of threatening jail sentences to the opposition. It has come to the point where he basically is saying that Zero votes from the opposition have been legitimate, and they have all been through bribes.
And just like that, two days ago we saw the Anti-Corruption Committee also present a draft law to ban political parties. So where are we now? How would you evaluate the outcome of this election?
Hrant: Right. Pashinyan is doubling down on divisive tone, and what he is opting for now is capture of the state. I mean, he’s setting the agenda of capturing of the state. And he did not get very good results from this election, even though this election was in many ways fabricated, but still.
And he is discontent with its results. And now he is trying to, you know, to somehow replace and to remove all his opponents. So on one hand, the elections were rigged. How they were rigged?
It was not the large-scale ballot stuffing, no. But it was an oppressive atmosphere. Hundreds of political figures were jailed and are still in jail. I mean, it was extremely oppressive.
On the other hand, it was very divisive and polarized election, and mainly it came from Pashinyan himself. And then in the end, there was also a large scale foreign interference in Armenian elections .
And lately we learned that French intelligence has helped Pashinyan to silence the critics also Meta has banned many opposition channels which they marked as so-called misinformation spreading content and so on maybe some of them were but you know there were very many pro-Pashinyan misinformation channels both on YouTube and Facebook and they did not ban any of them and the amount and the scope of misinformation was outstanding, but no reaction on their side. And in the end, they robbed Prosperous Armenia of its ability to participate in parliament.
So, in the end, they cancelled the results on three precincts, because of which Prosperous Armenia fell below the 4% threshold, and that is why now they cannot participate in Parliament’s work. So this is very obvious in front of everyone. And the Electoral Code of Armenia demands that the Central Electoral Commission should appoint new elections for these three precincts. And then the head of Central Electoral Commission made a political statement, not a professional statement, not a technical statement, but a political statement.
He says that the voters might change their mind and de facto vote in favor of Prosperous Armenia. Of course, opposition will go to the Constitutional Court and there are some signs that this might work. But anyway, the attempt to rob Prosperous Armenia of their votes and their seats in the parliament is very obvious. That is the result of these elections. and opposition did not recognize the results of these elections and did not recognize the authority which was formed as a result of these elections.
Seven opposition parties have denied these authorities of legitimacy but there were some parties which, you know, stayed neutral and we cannot say that those are opposition parties and then there is a ruling party, which is pushing on the other side. They say, yes, these elections should be cancelled because the opposition should not get any vote. So here we are, as of now.
Hovik: Yeah. Tell me more about your outlook in the coming months and years, because if this status quo remains and Pashinyan’s threats to further arrest opposition members materializes, essentially are we seeing that even more repression, even more jailing of opposition members to the point that they can’t even, you know, organize themselves properly? Or is this language of threats and intimidation just that and he will not act on it? What is the main goal of Pashinyan in According to the West, he won the elections.
So why amplify things? Why sharpen the rhetoric even more if it’s not going to result in even more arrests and more physical impediments to the opposition?
Hrant: Because this election was not about Armenian politics and about opinion of Armenian voters. That is why the whole world was very intensively participating in these elections, including the United States, Brussels, Turkey, and Russia. And especially Brussels and the United States, because those were directly involved in any step. And there was a technical group connecting Brussels and German, which helped Pashinyan on every step he made.
So in reality, we can say that all of what is happening right now in Armenian internal politics is either dictated by Brussels or is approved by Brussels so what they want if they approve all these oppressive policies so if they declare about you know win of democracy in reality crushing democracy like there is no tomorrow in Armenia so they want first of all they want to build anti-Russian frontier in Armenia and as Macron said two years ago Armenia is a buffer state against Russia so they want to construct a buffer state and the second is to achieve a solution which will favor Azerbaijan and that Azerbaijan will ultimately open the southern route so-called TRIPPP and they will have direct connection to Central Asia that’s what Brussels wants so fight against Russia in the name of Armenia and connection to Central Asia.
So if the goals are that big, then what Armenian people think or how harshly Pashinyan will crush on opposition, that does not matter. Now, Azerbaijan demands constitutional referendum. And opposition said no. And obviously, without a constitutional majority, Pashinyan will not be able to set a referendum.
That is also important. And without the qualified majority, which is three-fifths, Pashinyan will have to consult with the opposition on many accounts and if Prosperous Armenia is in the parliament then he will have to rule at least partially on a consensual basis not as a dictator and obviously he wants to be a dictator that’s all his rhetoric is about and then he says he needs constitutional majority to have peace .
In reality, he needs constitutional majority to have a referendum and European Union says we don’t care what happens here we don’t care who is right and who is wrong but Azerbaijan says that for the next step we need you know this referendum on constitution and we need a new constitution so we need this and Pashinyan says okay as you wish and for that they are trying to crack down on opposition completely and there are even discussions at least among the authorities and, you know, pro-Western circles, I mean, the organizations funded by the West, that they need to cancel the elections and hold new elections and use administrative resource and propaganda even more and foreign interference, by the way.
So in that case, they are sure they will get even higher results because according to what we can assume from the available poll data and from or other data, we can see that the level of support towards the ruling party has increased during the campaign very quickly, especially after the European Political Community summit in Yerevan in the beginning of May. So last month and a half was very beneficial for authorities. And also the administrative pressure has worked. So they understand that they need to double down.
That’s what they are heading to.
Hovik: But are they able to actually, you know, when you’re saying they believe if they cancel the elections and have new ones, will they have any different results if they don’t ban the opposition party? Or is that part of the plan? Banning all the opposition parties? So they can’t organize and then having new elections without the opposition?
Hrant: Well, in reality they don’t need to ban all the opposition parties. Kocharyan as opposition is very beneficial to Pashinyan.
Of course, Kocharyan is an opposition and his bloc is an opposition, but Pashinyan has played this black and white game and the previous regime versus me this game for last eight years and he wants to stay there since Samvel Karapetyan and his strong Armenian party is completely out of the of this frame that’s why he needs to remove this party from the whole political landscape and after that Kocharyan will obviously not be able to replace the strong Armenian party and in that case Pashinyan will get much a bigger share of votes That’s what he wants to do So I think that is at least an option Now is he able to do that?
Of course there is no obvious procedure which will let him to do so But that is why he is arresting more and more members of the party To show that this party is not a party, it’s a criminal gang So we need to stop it somehow and so on So that’s where we are being headed to
Hovik: Yeah, so Samvel Karapetyan earlier today had a press conference where he said that there were some mistakes. For instance, I think that he said the decision not to consolidate the whole opposition and not run in the elections as a consolidated bloc, even though there was some consolidation, I think he acknowledged that maybe that was a mistake. Can you identify any other opposition mistakes that could have mitigated the results we have today?
Hrant: Well, there were many of them in reality, but obviously the political landscape was very difficult and, again, foreign interference was unprecedented. Regarding the mistakes of the opposition, non-unifying was an obvious mistake. Another mistake would be, you know, not countering the fake narrative of Russia versus the West in the election, like the authorities represent the West and the opposition represents Russia and so on and so forth. And another thing not to counter the so-called TRIPPP and peace narrative. and the opposition did not have a good counter-plan on so-called peace agenda.
They said, oh, we will have another negotiation which will be better, but you need to have a better plan if you are proposing the major change in politics. That is also important and also in many occasions where Pashinyan was attacking opposition and maybe some other parties, opposition had to react more quickly, more adequately, but But you know, in reality, against such a level of administrative resource use, you cannot do much.
Even opposition leader, Samvel Karapetyan himself, who was jailed for a long period, even now he is not free. under home arrest, but under home arrest is still, you know, holding all these, you know, meetings and so on. So I think that the whole situation and political landscape is not free, is not letting Armenia to have free and fair elections. That is why we have this result. And in reality, if we have new elections, we are not going to have another result unless we have completely changed landscape.
Asbed: If we had another election, do you expect it to be more pro-opposition or more pro-government given the current circumstances?
Hrant: Well, if the trends continue, it will be more pro-government because government was very successful at terrifying the voters. In reality, voters in provinces are susceptible towards propaganda and towards both personal threats and the collective threats. That is why he managed this campaign of fear, fear of war, fear of losing jobs, fear of whatever you can imagine. And in reality, to sum up For example, what type of propaganda he was using against the opposition.
The ruling party’s propaganda narrative included that if Samvel Karapetyan wins, he will drag Armenia into war with Ukraine and send 300,000 troops to Russia-Ukraine front. So you can imagine how miserable this argument is. But for people in provinces who don’t know anything about geopolitics, this might sound adequate, and fearful if they hear this every day and every other hour.
Asbed: Hrant, what’s the recourse for the opposition? We are where we are right now. Should the opposition accept their mandates? Should they actually apply to the Constitutional Court?
Should they go for street action to demonstrate and show the government that they are unhappy with the results? What’s your suggestion as to how they should proceed from here?
Hrant: Well, I think opposition should take mandates. Many radical supporters of opposition or many radical critics of Pashinyan voiced that opposition should not accept mandates and in reality take a radical stance against Pashinyan, which would in reality lead towards non-recognition of parliament and the government. But that would have one of two possible consequences. One is opposition parties which are left out of the parliament, slowly dying out because of, you know, less participation and so on, which is another likely scenario, and we have seen that in many countries before.
Another scenario, if opposition takes radical stance and, you know, creates a shadow government and parliament and so on and so forth, then Pashinyan would crack down forcefully. and jail all of the members who are participating in that and that is why I think that with this kind of oppressive regime it would not work so in reality I don’t think opposition could achieve anything with not accepting mandates and if you don’t accept mandates take a radical stance and you have some big plan in which a very small portion of it is not taking mandates Yes, but it’s not this case and I don’t see any you know kind of action plan which opposition can have for further several months, so I think passion will continue the oppressive policy, but will he ban the opposition parties or not that isn’t clear for me.
Asbed: So days after the election results, Hikmet Hajiyev was in Armenia, in my opinion, to remind the Armenian government, Pashinyan’s government, about the deliverables that they had agreed to. In fact, Azerbaijani media has a bunch of articles about expectations about the constitutional referendum and results from that referendum. Let’s not assume that a referendum will automatically mean a change of constitution here. Now that we have a result that puts Civil Contract at only three-fifths and not two-thirds in the parliament, these are not going to be easy deliverables for Pashinyan.
How are we going to proceed from here? Is this the crux of the matter? Is this the reason why Pashinyan’s side is, I won’t say panicked, but quite concerned about these outcomes and how to proceed from here?
Hrant: Yeah, well, he knows what he has done. to win this election we can compare this election to 2021 election and see that 2021 election for example was compared to this election was much better organized and had much less traces of fabrication and so on and interference both foreign and administrative although it was not perfect but it was much better So we can say that in 2021, Pashinyan won honestly, at least in most cases. Now he knows that he would not win without all of this, you know, oppression which he has conducted. So he is struggling, understanding that as of now he cannot win clear elections.
And he started started involving, you know, police and all other stuff into the election since 2025 Vagharshapat elections. After Gyumri elections, which he has lost in Vagharshapat, he started already, you know, using all the good old techniques, which previous governments also used to win the elections. Now, but with one big difference, that aside from all those techniques, you know, carousel voting and so on,
Hovik: he also used jailing opposition and also if I can just interject he also had the support of the intelligence agencies of some of the largest countries superpowers in the world I don’t think that has happened in previous Armenian elections no no
Hrant: there has been nothing like that in the scale of foreign interference but anyway he understand that Even with all that, he did not get 50% of the vote. And he is nervous, of course, about that. And he has, you know, good experience from his good partner Erdogan , who has also cancelled several elections and then called new elections whenever he did not like the results. And arrested the winners, for example, Imamoglu is jailed, who has won the election in Istanbul.
So that is kind of thing Pashinyan would like to do. and also a mayor of Gyumri is jailed because he won the election in Gyumri and in many other cities and towns have the same has happened so that’s right and aside from that aside from that we are going to have a major local election round in September and October this year so just just three to four months after this moment and if the opposition works properly, Pashinyan will will lose most of those elections. Which local election is that? Well, that includes Vanadzor and many other towns. Many, many towns.
Majority of towns, I would say.
Asbed: Can I also add one thing, Hrant? Yesterday’s news included that the Civil Contract is considering annulling the elections in Gyumri, which they lost because they think that it was falsified. Of course, everything they do not win is considered falsified and they want to redo it.
Hrant: yeah in reality they tried to falsify it it was not enough and that’s why they want a new election and now they got 50%, or 50 and a half percent, in Gyumri and they are hoping that they will be able to win in Gyumri again so that’s that’s what they are trying to do and I mean the election does not matter for them the only thing they want to have full power and the reason for that again as I said control and pursue this pro-Azerbaijani agenda and lead Armenia toward a referendum and now with all this landscape Pashinyan might not be able to do that even after this election because if after Constitutional Court decision there will be re-election in all those three precincts then Prosperous Armenia will be back to parliament And that will mean that, in very many occasions, Pashinyan will have to consult with the opposition and that’s not what he’s planning to do.
So ultimately opposition will start showing force and it will change political landscape in Armenia gradually. I mean, Pashinyan’s support is based is based on on fear he’s planting fear in population but if he faces strong contradiction in the parliament that fear planting will not work ultimately because Armenia does not have sufficient potential to be a fully consolidated autocracy so that’s why Pashinyan is so hysteric I mean if if we had oil for example maybe he would not have to to make so many statements and speeches, he would just ban it and done.
But we don’t have oil, we gather taxes from citizens, and that means that citizens will have some to say, at least something. So he is not able to consolidate authoritarianism, but he wants to do so. And I don’t know, frankly speaking, where it will go in the future, because Many things are unclear as of now, but what I can say is that obviously the situation is far from end because it was not about elections, it’s about geopolitical frontiers coming into Armenia and being part of Armenian political landscape.
So the same geopolitical frontier which is passing through Ukraine in the form of a war front will now be a Cold War front in Armenian politics. That’s what we are observing.
Asbed: Hrant, so let’s turn our attention a little bit to the Russian aspect of these elections. Of course, Russia’s share of foreign interference was essentially telling Armenia that it’s very unhappy about its Western pivot, and they applied some threats and effective sanctions against Armenian exports, including agricultural products. Now, in response to this, the EU’s Ursula von der Leyen announced that the EU is going to provide Armenia with a $50 million line of credit, and this would help its resilience, basically, the economic resilience, meaning to say that they would be able to buy themselves out of the trouble they are in.
Pashinyan in response said that several countries in the EU have already approved Armenian agricultural products and they can start exporting these products to their countries. Among these, we already know that, for example, the Netherlands, which is famous for its flower industry, is going to open up its largest specialized flower marketplace to Armenian products. So far as I can see, the only real measure that we’ve seen so far is that Pashinyan is going to apply these lines of credit to subsidize the lost exports because European exports or rather European imports of Armenian products is not going to simply replace the Russian marketplace.
Hovik: Pashinyan joked that the Europeans will be buying so much Armenian products that, for example, France will buy so much apricots that Armenians in Armenia will call their relatives in Paris or Marseille and say, you know, when you come back to Armenia, can you bring some Armenian apricots with you? Because all of that will be consumed by Europe.
Hrant: Okay, let’s not reproduce his narrative We wanted to ask you,
Hovik: is all of this… We’re not, but how realistic is it? Essentially they’re painting that in one day they can flip the switch and Armenian products will go to Europe How realistic is that?
Hrant: Okay I think this situation is a very good test, not only for, you know, all this regional situation, Russia, Armenia, but also to the European Union itself. I mean, if this happened like 10 years ago, the Brussels response would be completely different from now. We see that Brussels is more and more getting involved into what is happening here. and also is starting helping the government of Armenia to counter Russian sanctions. That is very interesting because Russians have also sanctioned Georgia, Moldova in 2005 and Ukraine in 2014 and they did nothing.
And not only then, but even after Association Agreement with those countries, the European Union has done very little to help these countries improve their economic situation. I mean, they got the DCFTA, the Deep and Comprehensive Free Trade Area, but they got lots of quotas and so on.
And after that, these countries were not able to have extensive exports to the European Union but what we see now is that Brussels is reacting very quickly and they are doing something I would not say compensates completely a loss of Russian market but it will compensate for some share maybe I don’t know I don’t want to speculate but it will compensate somehow and that shows that Brussels is nowadays much more about geopolitics and less about economy, because economic lobbies across Brussels are very strong and even Polish economic lobby would push Brussels not to provide Ukraine with some access to market and so on and so forth.
But with Armenia, they are reacting very quickly. Why?
Because right now they need to counter the fire which has started with Russia banning Armenian products because it affects very very many families in Armenia and especially it affects Pashinyan’s electoral base which is mostly rural so they are trying to somehow react they are providing with some funds of course 50 million euros is enough only for one week but also they are giving some access to the market again logistics will be still constrained and then European standards and non-tariff regulations and so on and so forth, but at least something they have done, which they would not have done before. So this is also symptomatic change.
Asbed: Hovik, your question comes, it was very interesting because I was with some friends yesterday and we popped open two bottles, a bottle of French Armagnac and an Armenian brandy and tested both of them against each other, and both of them are excellent cognacs, basically, and it was very difficult. But if you were to ask me which one single bottle I would buy, it’d probably be the Armenian brandy. For me, it edged the French one for me.
Hovik: Yeah, but are the French going to prefer Armenian brandy over Hennessy or Courvoisier?
Asbed: Or the EU, that’s right. If there were like 100 million like me, maybe it’ll make a difference.
Hovik: In all honesty to me this just seems all like a psyop even the flowers I mean yes you mentioned Logistics I mean even if they open up the markets is how are you gonna? ship all that at a price that will be still competitive So I see the logistics as a big problem.
Asbed: So yeah, I mean everything has to be shipped by air basically and a long way
Hovik: Or by train, but by train it takes two months to get there. Good luck to your flowers. And there are reports that all the flowers in Armenia now are very cheap. So what’s happening is all these flowers are being sold in Armenian supermarkets at very cheap prices.
And I guess the government is just subsidizing the loss.
Asbed: That’s right. So that’s what’s going on. But more seriously, Hrant, I want to ask about the Russian aspect. Putin has basically withheld his congratulations of Pashinyan after the elections.
This is quite uncharacteristic for a so-called strategic partner at this point. Plus, only a couple of days ago, Russia’s director of foreign intelligence, Sergey Naryshkin, said that Armenia’s elections were questionable. What’s the message here for Armenia? Do you see Putin finally congratulating Pashinyan and under what circumstances?
Hrant: Well, I think the final decision on the Russian side will be accepting the political result without recognizing the elections. So they will say, we will deal with Armenian leadership, but we are not recognizing the elections. This formula is widespread in the West because they don’t recognize the majority of elections in developing countries. And for good reason, by the way.
And by the way, for more or less the same reason which we have here and now in Armenia. But anyway, I think that Russians will not provide Pashinyan with full recognition of the results of elections. Rather, they will say, okay, we know you are effective ruler of Armenia, that’s how we deal with you.
Asbed: But what would be the effect of not full recognition? What would be the difference between full recognition and the recognition that basically is Okay, we’ll just deal with you because you’re there.
Hrant: Well, there is some difference. And the difference is that in reality, Russia has almost never dealt with the opposition in any country. Not completely never, but almost never. And many political scientists in Russia have criticized the Russian authorities for that because they have only dealt with the government.
When a U.S. representative arrives to the country, Usually, they meet with the government and with the opposition as well. Except for Armenia, they don’t meet opposition here. But if they arrive to Georgia, they will meet opposition. This means that Russia will start working with the opposition too.
And that’s one thing. Of course, Pashinyan’s propaganda will use this as an excuse to call opposition pro-Russian. But I mean, They don’t have any real ground for that, so they will continue anyway Like you said,
Asbed: that is a fake narrative to talk about pro-Russian, pro-West in this situation That was just to polarize Armenian voter base Yeah,
Hrant: but anyway, I think that we will have more and more tension in Russian-Armenian relations And what will ultimately matter here is that many people in Pashinyan’s government would hope that after the elections, Russians would understand that, okay, the opposition lost, Pashinyan has won, so we need to deal with Pashinyan. And they will change the policy. And we see that Russians did not change the policy. They are increasing the number of economic pressure measures and so on.
So I think that here we see that, you know, the level of relations between Russia and Armenia just continues going down and down. So that’s a real outcome. It’s not something specific.
Hovik: Let’s move on to polling and talk a little bit about that topic, Hrant. After 2018, I think you mentioned this once or twice previously, polling in Armenia appears to be a very inaccurate predictor of actual voting results.
Now we regularly review polls and different polls and I’m wondering myself whether we have fallen ourselves in the trap of trying to use the polls as a proxy for predicting electoral results and now normally in a free and fair election we’ll be able to compare the election results and say okay here is where each poll went wrong but I don’t know if we even have that luxury because these results, as we said, have been completely, you know, manipulated in many different ways but I still think it’s worthwhile to look at the results so I put together a table from all the recent polls, whether it’s EVN Report, IRI, or MPG, and compared it with the actual results reported by the Central Electoral Commission.
And I’ve essentially tabulated the Civil Contract result and the main opposition result, with the main opposition taking into account Strong Armenia , Prosperous Armenia, Hayastan Dashinq , and I also included Wings of Unity for those polls where they were listed separately. Not all of them listed Wings of Unity separately. And I tried to estimate essentially the ratio of the regime versus the opposition in all these polls and I calculate two numbers so I’ll present these in the show notes for those but the two numbers are forecast accuracy score which is basically a root-mean-square error and then also balance accuracy score.
So forecast accuracy score measures how close a poll’s predicted vote shares were to the actual election results for both Civil Contract and the main opposition. So the lower the forecast accuracy score, the better. The same with balance accuracy score. It measures how accurately a poll captured the overall balance between the government and the opposition.
And I think I shared these numbers with you, but it was pretty interesting. In terms of forecast accuracy score, now I should begin by saying none of the polls came close, came even close to predicting Civil Contract’s final result. Every poll underestimated. And my Two metrics that I calculated found that the lowest or the best forecast accuracy score was from MPG, which was basically 14.3%.
That’s the error, essentially. So the lower, you know, if you had perfect prediction, it would be zero. For MPG, it was 14.3, and the highest was IRI, 24%. And then in terms of balance accuracy score, MPG was the worst and the best was an EVN Report in March, not the latest one, but towards the end of March.
So what do you think of this whole issue of using polls as a predictor of electoral results? Is it still a valuable exercise? And why do you think different pollsters produce such different estimates of the both the opposition support and Civil Contract support and none of them came even close to calculating Civil Contract’s actual results or estimating Civil Contract’s actual results.
Hrant: Okay, Hovik, you are trying to make sense out of very noisy data and your strong effort is whatever you do with this garbage data will not improve it.
Hovik: As they say, garbage in, garbage out.
Hrant: That’s exactly what I was trying to say. However you analyze this data, all of publicly available polls are problematic. Here we are. First, publicly available poll has someone who ordered it paid for it and paid for dissemination.
And if we don’t assume that someone was a good fairy and just wanted Armenian people to know about the state of public opinion, unless we believe that, we should assume that the motive was to somehow affect the public opinion. Not to inform public opinion, but to affect public opinion. Second, the timing is also very important because, okay, if you take the EVN Report poll from March, it might be close to the current result, but in reality it was very distorted on March because Pashinyan’s results had grown significantly since then, especially after May summit of, you know, the European Political Community,
Asbed: yeah.
Hrant: yeah and then the so called military parade one week before election and so on and so forth so I think that all of these polls were very distorted very distorted and given the very arbitrary circumstances like date like rate of responses You know, because majority of people have rejected participating in this vote and so on. You could get something which is closer to actual election result or which is far away. But since the election was also fabricated in many ways. I mean, it was not fabricated in a way that there was ballot stuffing.
But of course, there was extensive use of administrative resources, pressure on voters, and also around 100,000 of voters were brought to the precincts by the organizers of elections. For example, elderly voters who could not vote by themselves.
The representatives of a ruling party would take them and bring to the election, then enter the booth, vote instead of them and leave that’s exactly what Republicans would do in 2017 and now the ruling party is doing the same from the result table we can assume that up to 100,000 people like that were there so the real result will be somewhere close to 45% without all these you know actions but again if you rule out of all this foreign interference propaganda pressure on voters You don’t know what the result would be.
And that is why the opinion polling here does not work as well because whatever public opinion is, it is not transferred into election results.
Hovik: Yeah I think I think we were able to we have highlighted I consider like the non-response rates to be an indicator of people fearing talking and expressing their opinion but also even among the people who initially agreed to talk to the pollster the hidden vote where they said okay yes I agree to talk to you but then somehow, when they were asked more sensitive questions, they refused to respond. That was also very high. So not only is the initial polling response rate, but also the hidden vote extremely high. So does it even make sense for us to talk about polls in this way?
I mean, if they have no value, no predictive value, should we cancel our discussions on polling from now on? I would not say so.
Hrant: First, which party are you going to vote for is a very harsh question in this atmosphere which we were discussing last hour. In reality, the majority of other questions are not that much affected by the response rate. because what response rate does to the poll results? Those people who participate in polls and those who refuse to participate are different people and they have different opinions. Yeah, but they have different opinions based on what?
Based on the party they are going to vote for or based on, for example, based on foreign political preferences or maybe cultural preferences.
The less divisive the issue is, the less likely you are going to get distorted results of the polls So as of now, polls on who you are going to vote for are problematic, obviously But the polls regarding many other issues are still valid And regarding the election prediction and the polls I would say you still can somehow predict the results of the elections but you should not only take into consideration the result of polls but you need lots of metadata collected through the poll and you need very good quality polls and those are usually not published that is another issue there were many many polls that were conducted but not published so these polls are you know not accessible to the wider public but usually better in quality
Asbed: Hrant, some of these unpublished polls that you’re talking about, were they closer in reality?
Hrant: They were closer. I would not say that they had much. No, they did not. But, you know, for example, when you get refusals, you can ask, you refuse, but you are overall supporter of government or opposition.
And then you can get some responses at least some portion and then you can understand the coefficient which you need to use here and there then you need to have all of different sorts of metadata for the poll and then you are able to somehow reconstruct the real opinion of the wider public which they sometimes try to hide but then another question is that if public is in fear, then there is not much sense in measuring this opinion because it’s not transferred into election result. So that’s another end of the problem. So you have problems on both input and output. Input as a poll and output as election result.
Hovik: You mentioned that there is there are no polls in Armenia that are funded by fairies who have good intentions. But I’m just wondering, is this an opportunity for maybe a pro-Armenian organization, maybe, I don’t know, a groong? to, you know, if we find some fairy who can fund us the money, do you think that this would be something we can work with pollsters in Armenia to conduct open public polls with, like, all the relevant, you know, methodology done correctly?
And of course our goal is, you know, to be very accurate, so we would collaborate with, you know, we would try to find scientists to collaborate with Do you think that that could be something that’s useful or is it missing in the market today?
Hrant: I think you can try because at least there is a very low level of trust in polls and overall among the public towards any type of political content and you can try to at least have some regularity, have some topics covered and you can also try to do something different from what is usually done because you know you have different pollsters which have regular polls and they have some results which are then questionable but they have standard set of questions which are not that interesting and so on and they are as I said paid by some sponsors and so on now if you can set it differently that might have a good outcome theoretically.
Asbed: Hrant, when you say that there’s a very low level of trust in these polls, is that lack of trust fueled by the final results which are not very close to predicting what is going to become or is it based on, for example, the identity of of the polling itself. For example, the IRI poll is known to be a Western poll, so it certainly fuels a certain level of response and also non-response rate. Same with the MPG, same with the EVN Report, same with the Kopalyans and what have you. So how much of an effect does the perceived camp of the polling outfit make?
Hrant: Well, in reality, the pollster should not underline this during the interview. I mean, during the sociological interview, yeah, people might ask, who are you conducting this poll for? But usually you just say organization. It does not affect the surveyed people that much.
What affects the results more is a way of data processing,
Hovik: you know,
Hrant: because data might be processed differently. and sometimes it’s processed in a way which distorts the results in the direction which the people who are doing it they want to have this type of result that’s it and that is why I think that the problem is on the side of analysis and it almost always is because the analysis here is key to the results The raw data of results, you know, in Armenian public politics, majority of politicians, and not only politicians, but also other figures and civil society representatives, they just consider that the raw number is whatever is important. In reality, it is not. The interpretation is important.
And the processing is important because then you need to do weighting and so on and so forth. So interpretation is more important than raw data. That’s I can be sure for. And that is why we get distorted results because I think that raw data from all these polls is not very much different.
Asbed: So why would something, a poll like the IRI have a gigantic non-response rate? Well, I can’t remember right now. It was like 80 something percent basically hung up on them. Yeah, 84% just basically did not respond to the poll.
Hrant: Well, in reality, it’s not that unusual as of now. Maybe you remember the previous polls in Armenia, but some things have changed.
First, when sociology was starting in Armenia and the opinion polls were very new to population, people would love to respond, to talk to pollsters, but the situation has changed. that’s one and very important thing so the refusal rate is high lately during all the polls and it might be 75% might be 85% but not less than 75% and not more than 85% that’s the interval which the pollsters are dealing with right now that’s the reality so part of people just don’t want to participate especially those who are employed and who are in a hurry and who have some other things to do and so on that’s one part second part is political polarization of course which might be not that big share in reality but significant enough to distort the results that is also important because yes it obviously distorts the results and in case of IRI the refusal rate was like 84% so 16% was response rate but the thing is that other pollsters just don’t disclose the number That is very important because we just don’t know.
But from what I know, it’s usually between 15% and 25% is response rate, not more than 25%.
Asbed: Okay. Thank you, Hrant, for joining us today. This was a very interesting conversation. We’ll be in touch again sometime soon.
Hovik: Thank you, Hrant.
Asbed: That’s our show. We recorded the show on June 18, 2026. We’ve been talking with Hrant Mikaelian, who is a political scientist and multidisciplinary researcher in social sciences based in Yerevan. A very interesting fellow.
We are glad to have him on this show. Hovik, that was an interesting question that you raised. You and I have been talking a little bit about, you know, creating and conducting a Groong poll. What’s on your mind for our listeners?
Hovik: I mean, we have been talking about it, but what I found is that is a relatively expensive operation. but maybe through collaboration with like US universities and with data scientists and sociologists and political scientists both in Armenia and the US something could be created that is transparent that is regular that is done with full scientific standards where we publish our non-response rate Where we try to actually analyze More nuanced questions All of this without being funded By any foreign government The CIA, the Mossad The FSB Or the neocons I don’t know if it’s even possible It’s a tall order And given the amount of funding we receive Which is just enough to cover our hosting bills I don’t know if it’s even possible But if there is some way forward.
Asbed: Let’s be more specific. For example, if we are going to conduct a poll, just one poll, how much would that be? We’re talking about how much that we would have to crowdfund or find a way of funding. And I don’t even like the idea of a benefactor or a single benefactor because then suddenly people can say, oh, that benefactor is something or another.
Hovik: Yeah, we can probably do a Kickstarter and collect enough money, but that would require major campaigning. It’s about $20,000 or so? In terms of costs, I’ve talked to different pollsters, people who do this regularly, and of course I think that if we had to do an entire poll from scratch, it would cost anywhere from upwards of $10,000 per poll.
We can piggyback on other polls that are being conducted at the same time but we will probably require a regular cadence that is not tied to specific political events that you can try to filter out like the influence of one event during a year but yeah so the price tag for it is for a single poll all done for you like organized for you basically you have to imagine You have to call enough people by phone so that 1,000 people respond. Assuming a response rate of 20%, right? That would mean that… What’s a reasonable response rate?
Asbed: 30%, 40%?
Hovik: I think even 30% is too high. The response rate across polls and globally has tended to go down. So I think 20%, 25%, 30% is the best we can hope for. So if you want to reach,
Asbed: let’s say… You’re talking about 30% response rate or a non-response rate? 30% response rate. Ah, okay. So we almost have to triple the number that need to be called.
So you need to call almost 3,000 people to get a thousand people to respond.
Hovik: You get a response. Are they in the right demographic categories? So you have to actually call more people to ensure that all of your demographic groups are completed there is a lot of other things a lot of ways that these polls can be affected we talked a lot about them in the past you have to have a protocol for calling back how many times do you call back you have to ensure the same protocol for every respondent so maybe if they’re not if they’re busy then you say, okay, can I call back tomorrow at this time or something? So it’s a very technical process and there are companies and there are organizations in Armenia that can do this.
It’s just a very expensive process. So imagine every phone call taking 10 minutes, multiply that by probably around 5,000 calls to get like 1000 responses that are sufficiently weighted in different, you know, so you can do the math. It’s a manual effort.
Asbed: Okay, well folks, we do have plans, and they do cost money, so don’t forget to support us, podcasts.groong.org/donate. We would love to actually start piling up a little bit of money so that we can bring these ideas to fruition. Thanks for all the support. Thanks for listening to us.
I’m Asbed Bedrossian.
Hovik: I’m Hovik Manucharyan.
Asbed: We’ll talk to you very soon.
Hovik: Have an excellent week, and we’ll talk to you soon.