Experimental Evidence on Election Outcomes and Charitable Giving

Last registered on October 22, 2018

View Trial History

Pre-Trial

Trial Information

General Information

Title

Experimental Evidence on Election Outcomes and Charitable Giving

RCT ID

AEARCTR-0003472

Initial registration date

October 20, 2018

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

October 22, 2018, 1:04 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country

United States of America

Region

Primary Investigator

Name

Mattie Toma

Affiliation

University of Warwick

Contact Primary Investigator

Other Primary Investigator(s)

Additional Trial Information

Status

In development

Start date

2018-10-24

End date

2018-11-06

Keywords

Electoral, Welfare

Additional Keywords

JEL code(s)

D64, D72

Secondary IDs

Abstract

In the weeks after the 2016 presidential elections, support for left-leaning charitable organizations skyrocketed: In the first six weeks Planned Parenthood received donations at 40 times the normal rate, and by the six-month mark ACLU membership had more than quadrupled. Was this a one-time response to a uniquely controversial election, or is this suggestive of typical patterns of support for politically-bent charities? I aim to address this question by first conducting an observational data analysis to explore whether individuals are more generally inclined to give more to left-or right-leaning causes after the party with which they are ideologically affiliated loses power. I also seek to explore the possible mechanisms underlying such a trend, including (1) the emotional desire for retribution; (2) the increased salience of given causes around an election; or (3) a ‘crowd in’ effect where private donations increase when government support decreases. I plan to run an Amazon Mechanical Turk (MTurk) experiment around the midterm elections in order to both provide individual-level evidence for the more general trend of increased giving following political losses and also to specifically isolate the ‘retribution’ mechanism.

External Link(s)

Registration Citation

Citation

Toma, Mattie. 2018. "Experimental Evidence on Election Outcomes and Charitable Giving." AEA RCT Registry. October 22. https://doi.org/10.1257/rct.3472-1.0

Former Citation

Toma, Mattie. 2018. "Experimental Evidence on Election Outcomes and Charitable Giving." AEA RCT Registry. October 22. https://www.socialscienceregistry.org/trials/3472/history/36033

Sponsors & Partners

Experimental Details

Interventions

Intervention(s)

I will run the following cross-randomized interventions:

(1) News Treatment - Subjects will receive either good or bad news about the 2018 midterm elections, i.e. news suggesting that their own party is likely to do well or news suggesting the converse;

(2) Emotion Treatment - Subjects will be prompted to write about either how this news made them feel or what their beliefs are about votes in the upcoming elections.

Intervention Start Date

2018-10-24

Intervention End Date

2018-11-06

Primary Outcomes

Primary Outcomes (end points)

First, I am interested in whether subjects are more likely to indicate that they prefer to donate to the organization aligned with their political affiliation compared to the "neutral" organization when they receive bad news about the upcoming elections, and whether this effect is magnified when their emotions about this news are primed. Second, I am interested in whether donation amounts to a given organization increase under the same conditions.

Primary Outcomes (explanation)

Donation amount = the amount donated to a given organization by any one individual, where this amount is equal to zero when the individual selects an organization other than the organization in question.

Secondary Outcomes

Secondary Outcomes (end points)

Notable additional questions I plan to explore: (1) Does the strength of party affiliation affect the impact of my treatments?; (2) Are beliefs about the impact of the party in power on government support of causes the subject cares about correlated with giving decisions?; (3) What is my first stage effect, i.e. the impact of the good or bad news on beliefs about the upcoming elections?

Secondary Outcomes (explanation)

Experimental Design

I plan to administer a survey to MTurk participants in the two weeks before the 2018 midterm elections. As described above, the design will incorporate cross-randomized interventions, the "News Treatment" and the "Emotion Treatment." All subjects will then be asked which one of three nonprofit organizations they would most to donate: a Republican-leaning organization, a Democrat-leaning organization, or a "neutral" organization. After making their organization choice they will choose how much to donate to the organization versus keep for themselves as an additional study bonus. Donation decisions are incentivized: decisions made by one out of every ten participants are actually implemented. Finally, subjects will answer a series of questions about their backgrounds and political perspectives.

Experimental Design Details

Randomization Method

Randomization will be performed via Qualtrics, the survey platform used for this experiment.

Randomization Unit

Individuals will be cross-randomized to one of the two news treatments and one of the two emotion treatments.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

I plan to run this survey over 1000 participants in addition to ~100 pilot participants (pending pilot data indicating that a different sample size is needed to achieve my MDE, in which case I will update this registry accordingly).

Sample size: planned number of observations

1000

Sample size (or number of clusters) by treatment arms

1. Good News + Neutral Emotions Treatment - 250 subjects
2. Good News + Emotion Priming Treatment - 250 subjects
3. Bad News + Neutral Emotions Treatment - 250 subjects
4. Bad News + Emotion Priming Treatment - 250 subjects

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

Supporting Documents and Materials

IRB