Why do the polls decline? Analysis at historical and individual level

Response rate in recent years

Previous years have represented a "golden age" for opinion polls.
People were generally more likely to respond to the proposed questionnaires and response rates higher than 70% were not unusual.
In the last few years, the general participation in the surveys is much lower.

A study (Presser and McCulloch 2011) has shown the correlation between the huge increase in survey use in recent years and the corresponding decline in response rates.
In addition, the general trust in institutions (state, government, etc.) has precipitately declined in recent years, helping to distrust proposals and to increasingly decline the polls.

Subjective response rate

Subjectively, in making a decision, we calculate the gain-cost ratio.
With these terms is meant all that is classifiable as a gain from the respondent (the ability to count their opinion, to receive a monetary incentive ...) or as a cost (loss of time, etc.).
Recent studies show that the most frequent reasons that lead to accepting a survey are altruistic (contributing to the research objectives), selfish (redeem the incentive at the end of the compilation) or characteristic of the survey (interest regarding the topic dealt with, will to express and assert your opinion).
On the contrary, the reasons for rejecting a survey are the lack of interest in the subject, the lack of time, the presence of an incentive not commensurate with the commitment and reasons for privacy (fear that the personal opinion expressed could be in some way disclosed).


In order to obtain higher response rates, it is possible to experimentally manipulate the benefits and costs of a questionnaire.

The biggest variation in the response rate is obtained by increasing the respondent's gain or benefit, rather than by variations in the cost.
The topic of the questionnaire and personal interest in the questions have a great effect on the participation rate. In general, respondents are more likely to take part in questionnaires concerning work, hobbies or free time than in surveys about sex, money or income. "Sensitive" topics increase the perception of the "cost" of a survey.
Finally, in particular questionnaires where the compilation is mandatory or agreed with the subject, the warning of the penalties that may be incurred in the event of failure to complete and return the questionnaire greatly influence the response rate, to an even greater extent than increasing the “ I earn".


Increasing the benefits or decreasing the costs of a particular survey and measuring changes in the response rate in tests prior to the start of the survey, performed on a smaller number of respondents implies higher costs, but leads to higher response rates. This is particularly effective if the survey has as its primary importance the optimization of the response rate.

Eleanor Singer

Leave a comment