This dataset by the Electoral Integrity Project evaluates the quality of elections held around the world. Based on a rolling survey collecting the views of election experts, this research provides independent and reliable evidence to compare whether countries meet international standards of electoral integrity.
PEI-10.0 cumulative release covers 586 national parliamentary and presidential contests held worldwide in 170 countries from 1 July 2012 to 20 December 2023. For each contest, approximately 40 election experts receive an electronic invitation to fill the survey. The survey includes assessments from 5,230 election experts, with a 2023 response rate of 13%. The study presents two additive indices, collecting 47 indicators to compare elections for 2012-2022 and 62 indicators for 2023. The 2012-2022 indicators are clustered to evaluate eleven stages in the electoral cycle as well as generating an overall summary Perception of Electoral Integrity (PEI) 100-point index and comparative ranking. Starting in 2023, electoral integrity is measured through the aggregation of four concept indices (Contestation, Participation, Deliberation, and Adjudication).
Please note that for the QoG Data compilations, only three indicators are included: "Perception of Electoral Integrity Index", "Perception of Electoral Integrity Index Type" and "Electoral Integrity Rating".
Last updated by source: 2024-07-19
Dataset type: | Time-Series |
Dataset level: | Country |
(Garnett et al.,
2024)
The question below is answered in reference to the first election in the given county-year: Overall, how would you rate the integrity of this election on a scale from 1 (very poor) to 10 (very good)?
More about this variableThe question below is answered in reference to the second election in the given county-year: Overall, how would you rate the integrity of this election on a scale from 1 (very poor) to 10 (very good)?
More about this variableThe question below is answered in reference to the third election in the given county-year: Overall, how would you rate the integrity of this election on a scale from 1 (very poor) to 10 (very good)?
More about this variableThe PEI index is designed to provide an overall summary evaluation of expert perceptions that an election meets international standards and global norms. It is generated at the individual level using experts' answers to the 47 substantive variables. Mean substitution is used at the expert level to fill missing data. The 47 scores are summed and then standardized on a 100-point scale. Starting from PEI 10.0, they reduced the number of questions included in this calculation from 49 to 47. The two deleted variables are postal voting and internet voting availability. These were removed since authors do not believe them to be essential to electoral integrity, since a variety of convenience voting methods are available in countries.
More about this variableElectoral Integrity Index, calculated for the second election in a year. The PEI index is designed to provide an overall summary evaluation of expert perceptions that an election meets international standards and global norms. It is generated at the individual level using experts' answers to the 47 substantive variables. Mean substitution is used at the expert level to fill missing data. The 47 scores are summed and then standardized on a 100-point scale. Starting from PEI 10.0, they reduced the number of questions included in this calculation from 49 to 47. The two deleted variables are postal voting and internet voting availability. These were removed since authors do not believe them to be essential to electoral integrity, since a variety of convenience voting methods are available in countries.
More about this variableElectoral Integrity Index, calculated for the third election in a year. The PEI index is designed to provide an overall summary evaluation of expert perceptions that an election meets international standards and global norms. It is generated at the individual level using experts' answers to the 47 substantive variables. Mean substitution is used at the expert level to fill missing data. The 47 scores are summed and then standardized on a 100-point scale. Starting from PEI 10.0, they reduced the number of questions included in this calculation from 49 to 47. The two deleted variables are postal voting and internet voting availability. These were removed since authors do not believe them to be essential to electoral integrity, since a variety of convenience voting methods are available in countries.
More about this variableClassification of the PEI Index on five categories. 1. Very Low 2. Low 3. Moderate 4. High 5. Very High
More about this variable