This paper investigates potential biases in citizen-based corruption indicators.
How Respondents Answer
- Standard questions may be subject to political bias, where answers reflect partisan identity rather than objective perceptions. This is tested using priming experiments that manipulate respondents' political affiliations before asking about corruption.
- Corruption perception questions are also affected by sensitivity bias, which means people might not report experiences honestly due to potential stigma or fear of repercussions. A list experiment technique was used to mitigate this issue.
What the Findings Reveal
- Significant response biases exist in standard survey methods for measuring corruption perceptions and experiences.
- These biases are not uniform across populations but vary substantially between different subgroups.
Why This Matters Now
- The study highlights that widely-used corruption indicators may be systematically distorted.
- Political scientists should adopt more sophisticated measurement approaches when studying public attitudes toward corruption.
