Effective mechanisms to resolve resource allocation problems

How to cite

Rusch, Hannes; Morath, Florian; Glowacki, Luke. 2019. "Effective mechanisms to resolve resource allocation problems." AEA RCT Registry. AEARCTR-0004847. https://doi.org/10.1257/rct.4847-1.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Hannes Rusch* , Florian Morath , Luke Glowacki

* Primary investigator

Registry ID
AEARCTR-0004847
DOI
https://doi.org/10.1257/rct.4847-1.1
First registered
Last updated
Status
Completed
Trial dates
2019-10-14 – 2021-12-23
Countries
Ethiopia (Southern Nations, Nationalities, and Peoples' Region)
JEL codes
D72, D74, C92, C93
Registry record
http://www.socialscienceregistry.org/trials/4847

Abstract

We analyze factors that explain the success of democratic institutions as a means to solve resource allocation problems. Our lab-in-the-field experiment investigates the following questions: Can democratic allocation mechanisms effectively resolve distributional conflict in the presence of ancient hostilities between groups? Does the risk of being exploited by the majority induce the minority to reject democratic allocation institutions and trigger inefficient conflict? Do restrictions of the majority's decision-making power resolve the distributional conflict and increase the success of democratic allocation institutions?
Our experiment is based on a game theoretic model in which two groups of unequal size decide whether to implement a given democratic allocation mechanism or to choose conflict before they observe the outcome of the allocation mechanism. Our main experimental variations address the impact of ethnic hostilities on the likelihood of conflict as well as the role of the--explicit and implicit--decision-making power of the majority. For the former, we compare the success of democratic institutions in the presence versus absence of ancient hostilities. For the latter, we compare two variants that make the political institutions more ‘inclusive’: (a) a change in the voting rule that strengthens the minority in the democratic allocation mechanism; and (b) the threat of triggering conflict as an implicit option to exert veto power in case of too unequal democratic resource allocations.

Intervention

Our baseline treatment (“MAJ-DIFF”) implements majority voting as the mechanism to decide on the allocation in case the groups do not opt for conflict. The two groups are recruited from different ethnicities. As the first variation, the same game is played among individuals from the same ethnic group (treatment “MAJ-SAME”). As second and third manipulations, we change (a) the voting rule (treatment “PR-DIFF”) and (b) the timing of the game (treatment “MAJ-VETO”). In PR-DIFF we give each individual the same chance to decide the allocation outcome (as with a random dictator rule). In MAJ-VETO we change the sequence of actions, such that individuals can decide on conflict conditional on the outcome of the democratic allocation institution.

Experimental design

Our baseline treatment (“MAJ-DIFF”) implements majority voting as the mechanism to decide on the allocation in case the groups do not opt for conflict. The two groups are recruited from different ethnicities. As the first variation, the same game is played among individuals from the same ethnic group (treatment “MAJ-SAME”). As second and third manipulations, we change (a) the voting rule (treatment “PR-DIFF”) and (b) the timing of the game (treatment “MAJ-VETO”). In PR-DIFF we give each individual the same chance to decide the allocation outcome (as with a random dictator rule). In MAJ-VETO we change the sequence of actions, such that individuals can decide on conflict conditional on the outcome of the democratic allocation institution.

Design details

Randomization method
Participants will be randomly assigned to sessions in a walk-into-the-lab manner. Sessions of different treatments will be randomized over the study’s data collection time frame.
Randomization unit
Individual: participants are randomly assigned to a treatment, i.e. the treatment used in that particular session that they happen to take part in. Note that participants will be selected based on gender and their ethnicity: we only recruit male participants from two major ethnic groups residing in the study region and we form ethnically homogenous experimental sessions. We only recruit male participants in order to avoid gender effects when our participants interact within groups and have to make joint decisions. Ethiopia is a patriarchal country in which women traditionally are often excluded from democratic decision-making processes.
Clusters
Our treatments are clustered in experimental sessions: We plan on running up to 10 sessions per treatment, each with either 6 or 12 participants.
Observations
240 adult male participants
Arms
60 participants in between 4 and 10 experimental sessions per treatment
Primary outcomes
Our main variable of interest is the likelihood of a choice of conflict in the four treatments. As secondary outcome variables, we compare (i) the allocation proposals under the different treatment conditions, (ii) individuals' beliefs about the choices of others, and (iii) the aggregation of individual choices to a joint group decision. Our main predictions are derived as the subgame-perfect Nash equilibrium of a standard game theoretic two-stage model of conflict which allows for in-group altruism and out-group spite.

Cited by

No citation data available for this registration yet.

Keywords