press release [imas]27 february 2019
survey results versus election results
The elections held on 24 February, 2019 were carried out based on the mixed voting system, and this implied expressing the vote based on the proportional system (on party list) and the vote for the candidates registered for single-mandate constituencies.
Different types of analyses and conclusions are spread in the public space regarding the accuracy of sociological surveys. We would like to clarify things in this respect:
- It is incorrect to compare the sociological surveys’ results with the results regarding the number of seats obtained by every political party, as the national surveys tackled people’s intentions for the proportional vote (party lists) and not the voting intentions for single-mandate constituencies.
- It is incorrect to compare the results of the net voting intention with the results registered for party-list voting, as besides the people expressing their voting intention, every survey also presents the number of undecided people. These persons also have to be considered when making comparisons between the estimates provided by opinion polls and the results obtained for proportional voting (party lists).
- The correct comparison to be made would be between the results of the net intention to vote plus a certain share of the undecided people allocated per every party versus the results registered by political parties for the proportional voting system (vote on party lists), options expressed on the territory of the Republic of Moldova. The published opinion polls did not refer to voters from diaspora, but only the ones on the territory of the Republic of Moldova.
Please see below the results of the most recent survey [imas] – the net voting intention, the number of undecided people per every political entity (with their proportional distribution depending on the voting intention) versus the results registered on the territory of the Republic of Moldova according to the Central Electoral Commission.
We would like to present, as well, the results of the most recent polls published before elections:
Several conclusions may be easily drawn from the data presented in the above table:
- [imas] survey has estimated with the smallest margins the result registered during the elections
- [imas] survey is the only one which has estimated correctly the parties’ order in the ranking
- [imas] survey is the only one which has estimated that Sor Party will pass the electoral threshold
- except for [imas] survey, all the others have over-estimated PSRM’s rating, with SASD survey registering a deviation of 17.2%, although it was carried out very close to the elections’ date
- except for [imas] survey, all the others have under-estimated PDM’s rating, with International Republican Institute (Magenta) survey having the largest deviation – of 8.3%
- except for [imas] survey, all the others have under-estimated Sor Party rating, with SASD survey having the largest deviation – of 5.4%
- except for SASD survey, all the others had estimates close to the result registered by ACUM electoral bloc; the deviation registered by SASD survey in this case was 8.2%
- considering the average and maximum deviations registered by the surveys for the parties which passed the electoral threshold, the polls’ ranking would be as follows: [imas], iData, CBS-AXA, Magenta, and SASD
Of course, there is always room for improvement and we are aware of this. [imas] company will continue working with the same level of passion and honesty. I would like to thank the entire [imas] team – drivers, operators, controllers, and colleagues from the central office in Chisinau – their professionalism is undeniable, just like the above data show. And many thanks to the citizens who participate in our surveys, for their patience and time provided when taking part in opinion polls.
general manager [imas]