Presidential poll – part two


Aleksandar Krzalovski

Last fall I had a trilogy of columns devoted to polls. Some of them referred to the data / results of those surveys, but in most cases I tried to demystify some things about the survey process itself and what should be taken into account when interpreting various survey results. The size and representativeness of the sample, through the method of survey (telephone, field or other) and the formulation of questions and possible answers, to the statistical error and the credibility and professionalism of the agencies that examine the public opinion.

After the three candidates for the presidential elections were confirmed: Stevo Pendarovski, Gordana Siljanovska-Davkova and Blerim Reka, it is time to comment on the first survey on this topic, commissioned by MCIC and IDSCS, and conducted by the agency M-Prospect on a representative sample of 1.001 respondents, between 13th and 19th March. Before I comment on the results, let me point out two or three things that have caused confusion in the previous polls (and I expect some confusion with this one as well).

The first thing that comes to mind when comparing this poll with the previous presidential elections is the dramatic increase in the rating of the candidates (at least two who were “in circulation” and the previous polls). They are now in the range of 22-25 percent, and previously they were from 2 percent to 7 percent. The answer or “the trick” lies in the number of candidates offered by the survey, and in the question itself.

It is one thing to ask: “How much confidence do you have in the politician AB?” And quite another to ask: “In which of the abovementioned politicians do you have the most confidence?” In the first question, the poll asks about one person and in fact it is the real measure of “ratings” of that person, that is, the level of confidence of the citizens in him / her. However, such questions are rarely placed on the polls in our country (because of the price, among other things, because special questions should be asked for all possible persons for which we want to evaluate confidence, as opposed to the option to do this with just one question). In the second case, the choice is among many candidates and logically, because of simple mathematics, the “total rating” of 100 percent is shared among them, that is, the individual levels of confidence are much smaller than the real one (if only one person is asked). So, in these polls, when there were 10 candidates on the list, that is, the options for selecting / responding to the surveyed citizens, the individual estimates ranged between 1 and 10 percent, except in the case of Prime Minister Zaev (this is an additional problem when the leaders of political parties are on such lists, usually they receive the most attention and the highest percentage of votes of the respondents.
A special situation from the very way of asking the question is which answers will be offered to the respondents. It is quite common to ask the question “In which of the following politicians do you have the greatest confidence?”, and give the respondents a card with a list of names (in the field survey) or read the names from the list (telephone). It is usual to have an option “Other” and they themselves can indicate, but it is possible that the interviewer does not give that option. It is possible that the survey sometimes will not list specific people, but the respondents themselves can tell in which politician they have confidence. All these variations can significantly affect the individual ratings of each of the candidates, but basically show the current relationship of confidence between each of them, therefore more often this type of question is being practiced (for more individuals in one question).

All this affects the results and their definition, ie interpretation, without knowing all these things and “jumping” into hasty conclusions, can lead to a completely wrong image from what the results actually show.

The first impression is the equality of the two main candidates – Pendarovski (25 percent) and Siljanovska-Davkova (22.4 percent). The candidate supported by SDSM and DUI (and their coalition partners) is in a slight advantage (2.6 percentage points in the first round and 6.2 for the second one), but that is within the statistical error (3.1%), which means that cannot speak of a certain / clear advantage, and even less for a sure victory of some of the candidates.

It is certain, however, that exactly these two candidates will find themselves in the second round, since although Blerim Reka is the only ethnic Albanian candidate, he does not have enough potential to reach at least the second place after the number of votes in the first round (and because he is not supported by all ethnic Albanian parties, among other things).

His candidacy makes the race more interesting in terms of the behavior of the voters of Albanian ethnicity. While it is clear that the main body of votes for him will come from supporters of the parties that backed his candidacy, it is also worth mentioning that in the first round, more than half (57%) of DUI supporters would vote for him. The behavior of these (DUI) voters in the second round will turn to Pendarovski’s candidate and their support will grow from 25.7 percent in the first round to 58.6 percent in the second round. If it really happened so at the very elections (it’s early still too in the process and the presidential campaign has not even started yet), this is likely to be one of the decisive factors in winning the election.
This is the second important impression of the results that the VMRO-DPMNE’s candidate (and the party itself, above all) still fail to recover part of the electorate from the ethnic Albanians. The support is insignificant (below 1%) and in the first and second round, and although Siljanovska-Davkova has the advantage of the ethnic Macedonians (6.1 percent and 7.3 percent respectively for the two election rounds), it will be insufficient for victory, due to the number of votes Pendarovski will receive from ethnic Albanians. Of course, this will also depend on the turnout of these citizens – many of them will certainly not come out to vote, especially in the second round (among others, as a protesting voice towards the policies and behavior of DUI), since most of them (37.3%) stated that will not vote in the second round (as opposed to 34.8 percent of all ethnic Albanians who would vote for Pendarovski), which is more than double of those who will not vote at all (even in the first round) – 16.8 percent.
In the end, just a few words about the characteristics of the candidates. The survey shows the equality of questions of honesty, familiarity with the people and clarity / comprehensibility of what they say (in all these cases, the difference between Pendarovski and Siljanovska is less than 2 percent). Siljanovska is seen as a more independent (from political parties) candidate (24.3 percent versus 19 percent). On the other hand, Pendarovski has a greater reputation among foreigners (39.3 versus 14.2 percent).

I have two comments on this. First of all, no matter how hard the candidates, and the parties themselves, try to be presented as independent or non-party (or “consensual”), most of the citizens do not see them as such, and the very “partisan” distribution of the respondents’ answers continue to confirm the full polarization of the society on that (party) basis (for all candidates, over 85% of the supporters of the party that supported them will vote for them).
Secondly, although they all are good candidates, all PhDs and university professors, with long and respectable careers, the data is concerning (it should worry them as well) that the question “Which of the candidates do you consider the most honorable?”, Most of them of the citizens (34.2%) answered “None!” This only confirms to what extent the parties have led the political culture in the country, so everybody has to reconsider whether they are willing to deal with it – even if it about a presidential candidate!
However, there is an interesting campaign ahead, I hope it will be the most peaceful and most decent so far (although the ratings point to a close fight, which is usually  subject to abuse of the electoral process to get to victory). Of course, there are always individuals who, of course, “know everything” and already say that “everything is planned ahead” and that Pendarovski’s victory is certain (and agreed), but I am one of those naïve people and do not believe in any of the “conspiracy theories”, so I will enjoy the uncertainty that comes with each elections.

Views expressed in this article are personal views of the author and do not represent the editorial policy of Nezavisen Vesnik