Voting Secrecy Guaranteed

I would like to voice some reservations about Ukrainian society’s failure to discuss the method of anonymous polling, which was used in the exit poll in the first round of the October 31 presidential elections. The secret ballot method has been used in exit polls since the early 1990s, even though the earliest experiments with this method began in the 1940s. The problem of untruthful answers and refusals to participate in polls is one of the key factors that affect the reliability of exit poll data. We therefore proceed from the assumption that if the respondent feels that he will be guaranteed complete anonymity, he will be more honest in answering the proposed questions.

What is the secret ballot method all about? Instead of openly asking respondents whom they voted for, interviewers suggest that respondents mark their answers on a special ballot with the names of the presidential nominees and deposit it into a sealed cardboard box containing other respondents’ ballots. Results of many experiments suggest that secret ballots in exit polls more closely match the official vote count than the results of personal interviews. There are several reasons that would account for such differences.

First, if people think that their opinion places them in a social minority, they refrain from voicing it in public so as to avoid social ostracism.

Second, some potential respondents may not believe that the poll is anonymous and are afraid of voicing their electoral preferences in case they are subjected to certain sanctions.

Third, the interviewer may exert influence on the respondent, i.e., the interviewer’s age, sex, or appearance may discourage respondents from revealing their electoral preferences.

Taking into account the positive international experience of using secret ballots in exit polls, as well as interesting experiments with this method that have been conducted by Ukrainian scholars, we decided it was appropriate to test it out in a nationwide exit poll conducted by a consortium of independent public opinion firms. As part of the experiment, two public opinion companies conducted an exit poll by personally interviewing voters, while the other two used the secret ballot method.

We expected the secret ballots to somewhat reduce the number of people who refused to take part in the exit poll. Incidentally, during the exit poll of the 2002 parliamentary elections 24.5% of potential respondents refused to give answers. This year we expected this percentage to increase even more because of intense political strife.

According to estimates by various experts, the secret ballot method helps obtain data that significantly differ from the results obtained using other polling methods, primarily the method of personal interviews. The standard margin of error is two to three percentage points. As opposed to personal interviews, the results of secret ballots normally show more support for generally unpopular candidates in any given region, as well as candidates in opposition to those in power.

The history of this method is described in greater detail in:

Bishop, George F. and Fisher, Bonnie S. “Secret Ballots and Self-Reports in an Exit-Poll Experiment.” Public Opinion Quarterly, vol. 59 (1995): 568-88.

Perry, Paul, “Certain Problems in Election Survey Methodology.” Public Opinion Quarterly, vol. 43, no. 5 (1979).

The spiral of silence theory is described in greater detail in:

Benson, Lawrence E. “Studies in Secret-Ballot Technique.” Public Opinion Quarterly, vol. 5 (1941).

Bolshov, Ye. and N. Kharchenko. “Rezultaty eksperymentu iz zastosuvanniam tekhniky secret ballot u vyvchenni elektoralnykh skhylnostei ukrainskykh hromadian” [Results of an Experiment Using the Secret Ballot Technique to Study Electoral Leanings of Ukrainian Citizens]

Hoek, Janet, and Philip Gendall, “Factors Affecting Political Poll Accuracy: An Analysis of Undecided Respondents.” Marketing Bulletin, no. 8 (1997): 1-14.

McDonald, D., C. Glynn, and others. “The Spiral of Silence in the 1948 Presidential Election.” Communication Research, no. 2 (April 2001): 139-55.

Mishchenko, M. “Issledovanie vliianiia interviuera na otvety respondenta: rezultaty metodicheskogo eksperimenta” [Study of the Interviewer’s Influence on the Respondent’s Answers: Results of a Methodological Experiment]

Perry, Paul. “Certain Problems in Election Survey Methodology.” Public Opinion Quarterly, vol. 43, no. 5 (1979).