|
Field
Abstract
|
Before
Sufficient voter turnout is a desirable property of a healthy democratic system. Unfortunately, historical data show that a sizable portion of the eligible electorate tends to abstain. This project aims to increase voter turnout by leveraging important insights from personal and social psychology and behavioral economics.
The main idea is to modify the current vote counting rule in the following manner: uncast votes are not ignored but rather, randomly allocated to the alternatives. We hypothesize that this intervention will prompt people to turn out to vote to avoid having their votes misused or misallocated by the random mechanism. The intervention does not restrict voters' freedom (as opposed to, e.g., obligatory voting) and would be easy to implement in real life. We test the effectiveness of the proposed intervention both in the field (U.S. presidential elections) and in the laboratory experiments in Switzerland and Germany.
|
After
Sufficient voter turnout is a desirable property of a healthy democratic system. Unfortunately, historical data show that a sizable portion of the eligible electorate tends to abstain. This project aims to increase voter turnout by leveraging important insights from personal and social psychology and behavioral economics.
The main idea is to modify the current vote counting rule in the following manner: uncast votes are not ignored but rather, randomly allocated to the alternatives. We hypothesize that this intervention will prompt people to turn out to vote to avoid having their votes misused or misallocated by the random mechanism. The intervention does not restrict voters' freedom (as opposed to, e.g., obligatory voting) and would be easy to implement in real life. We test the effectiveness of the proposed intervention both in the field (U.S. presidential elections) and in the laboratory experiments in Switzerland.
|
|
Field
Last Published
|
Before
October 28, 2020 09:14 AM
|
After
October 19, 2021 06:37 AM
|
|
Field
Experimental Design (Public)
|
Before
We study several variations of randomization mechanism and include treatments with pecuniary penalty for abstaining.
|
After
We study several variations of randomization mechanism.
|
|
Field
Planned Number of Clusters
|
Before
1800 for the US. survey;
200 for the experiment in Germany;
200 for the experiment in Switzerland.
|
After
1800 for the US. survey;
600 for the experiment in Switzerland.
|
|
Field
Planned Number of Observations
|
Before
1800 for the US. survey;
200 for the experiment in Germany;
200 for the experiment in Switzerland.
|
After
1800 for the US. survey;
600 for the lab study in Switzerland.
|
|
Field
Sample size (or number of clusters) by treatment arms
|
Before
600 per condition for the U.S. survey;
40 per conditions in lab experiments.
|
After
600 per condition for the U.S. survey;
200 per condition (3 conditions in total) in the lab experiment.
|
|
Field
Keyword(s)
|
Before
Behavior, Electoral, Lab
|
After
Behavior, Electoral, Lab
|
|
Field
Intervention (Hidden)
|
Before
Instead of ignoring uncast votes, a common feature of election process, we propose to randomly allocate them among the eligible candidates. The randomization is either exercised ex-post (only the votes of those who did not cast their votes are subject to random allocation) or ex-ante (the votes of all eligible voters are randomly allocated among the candidates, voters who cast their votes overrule the initial allocation, while the votes of those who do not turn out to vote remain allocated to the candidate as per initial randomization). We vary the randomization procedures and compare the effectiveness of the intervention to a pecuniary penalty for abstaining from voting.
|
After
Instead of ignoring uncast votes, a common feature of election process, we propose to randomly allocate them among the eligible candidates. The randomization is either exercised ex-post (only the votes of those who did not cast their votes are subject to random allocation) or ex-ante (the votes of all eligible voters are randomly allocated among the candidates, voters who cast their votes overrule the initial allocation, while the votes of those who do not turn out to vote remain allocated to the candidate as per initial randomization). We vary the randomization procedures and compare participation rates with and without the intervention.
|
|
Field
Building on Existing Work
|
Before
|
After
No
|