Correlation Neglect in Voting Decisions: An Experiment

Moser, Johannes; Wallmeier, Niklas

doi:10.1016/j.econlet.2020.109656

All research

Journal article2021Expectations and forecasting

Correlation Neglect in Voting Decisions: An Experiment

Johannes Moser and Niklas Wallmeier

Economics Letters 198, 109656

Journal Cite

A schematic, not data: a voter weighs signals about which option pays off for the group against her own attitude to risk. Counting two signals from the same source as independent tips the balance; the vote flips, and in the experiment this left groups better off.

In short

An experiment on whether voters account for the correlation between the pieces of information they receive.

What we find

Members of a group vote for a fixed payment or for a lottery with the same expected value. Signals hint at which option may earn the group a bonus or a penalty.
Each voter has to weigh these signals against her own attitude to risk.
Correlation neglect, treating signals from a shared source as if they were independent, makes voters flip their vote.
The bias improves efficiency: the groups end up better off.

Cite

Johannes Moser and Niklas Wallmeier (2021). “Correlation Neglect in Voting Decisions: An Experiment.” Economics Letters 198, 109656. https://doi.org/10.1016/j.econlet.2020.109656

@article{moser2021correlation,
  author = {Moser, Johannes and Wallmeier, Niklas},
  title = {{Correlation Neglect in Voting Decisions: An Experiment}},
  year = {2021},
  journal = {Economics Letters},
  volume = {198},
  pages = {109656},
  doi = {10.1016/j.econlet.2020.109656}
}