It’s a small world: effects of overlapping neighborhoods in the provision of public goods

How to cite

Romero-Fernández, Lorenzo; Lange, Andreas. 2022. "It’s a small world: effects of overlapping neighborhoods in the provision of public goods." AEA RCT Registry. AEARCTR-0009451. https://doi.org/10.1257/rct.9451-2.1

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

Lorenzo Romero-Fernández* , Andreas Lange

* Primary investigator

Registry ID
AEARCTR-0009451
DOI
https://doi.org/10.1257/rct.9451-2.1
First registered
Last updated
Status
In development
Trial dates
2022-05-23 – 2022-07-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/9451

Abstract

This project studies the voluntary provision of a weakest-link public good in overlapping neighborhoods. Located around a virtual table, each individual’s investment decision affects the provision of the public good of herself and her right- and left-hand neighbors. We hypothesize that the introduction of this spatial element can provide obstacles to the level at which the public good is provided and can create a dynamic to the provision level across space. We compare treatments in which the endowments are homogeneous with those where some subjects receive a large, some a low endowment. For this heterogeneous case, we specifically study how the spatial distribution of different income levels affects the investment decisions. We investigate and compare two different mechanisms through which subjects can invest at neighbors’ locations, a direct in-kind transfer system to any neighbour vs. one where subjects can contribute to a common common that is divided up among low income participants.

Experimental design

This experiment will be conducted by the experimental laboratory of the University of Hamburg (Germany) in an online setting. The experiment will make use of a weakest-link public good game with 6 different treatments. In particular, participants will be assigned to groups of 6 players who interact in a 10-round public good game. The addition of overlapping neighborhoods extends the standard public good game in this study.
At the beginning of each of the 10 rounds, each participant is allotted again her initial income and the game is repeated. In other words, there is no possible transfer of tokens between rounds. Each plyer decides in each round how many tokens to invest in the public good in her own position, how many to invest by an in-kind transfer to any other location, and how many to keep in her account. At the end of each round, each participant will get an outcome given by the tokens that she has left minus the transfers that she made and the payback from the public good provided at her position. The public good has a unitary cost of 1 and each participant gets a benefit of 2.5 times the amount of public good provided at her place. The formula for the outcome for a player I is then: Outcome = Tokens remaining in own account MINUS Sum of Tokens invested in any location PLUS 2.5*Amount of total investment at her position.
In treatment 1, a group of 6 players is organised in 2 subgroups of 3 participants each. In this treatment, all participants have a homogeneous income of 30 tokens each. In each of these subgroups, each participant makes a decision on how many of these 30 tokens to invest in her own position and how many to transfer and to which groupmate. The investment in the public good are aggregated by a weakest-link technology, so the amount of public good provided to each participant in a subgroup is the minimum amount invested by any of them plus the potential transfers received from the others.
In treatment 2, a group of 6 players plays a public good game in a virtual circular neighbor. All 6 participants have a homogeneous income of 30 tokens. In this case, the provision of the public good in the position of a player will be driven by the minimum provided in the neighbourhood formed by her, and her right- and left-hand neighbors. This way of aggregating the investments will create 6 overlapping neighborhoods, one for each participant.
Treatments 3 and 4 have the same setting as treatment 2, but they have heterogeneous income for the participants – in each group, 3 participants will have a high income of 40 (H), and 3 will have a low income of 20 (L). In treatment 3, the participants are spatially distributed alternatively regarding their type (H-L-H-L-H-L), while in treatment 4 they are clustered together (H-H-H-L-L-L).
Treatment 5 and 6 have the same payoff structure as treatments 3 and 4, but investments can only be made at the own location of a participant and into a common chest. The total amount invested into the common chest works is divided up equally between the three low-endowment participants.

Design details

Randomization method
Randomization into treatments as well as locations and roles within the experiment done by computer.
Randomization unit
Groups of six
Observations
432 observations/participants (6 treatments, each having 12 groups of 6 participants). Thus, 12 independent observations per treatment.
Primary outcomes
We investigate the treatment effects on investment levels and transfer decisions. Hypothesis 1: The overlapping-neighborhood setting will decrease the average investment levels. For testing hypothesis 1, we will compare the investment levels in the treatments where there is no overlapping neighborhoods and where there is one with a homogeneous income distribution (treatment 1 vs. treatment 2). Hypothesis 2.a: For the heterogeneous-income treatments, the spatial distribution of the different types will affect the investment decisions – in the distribution where the high- and low-income type are alternated the investment will be larger than where they are clustered. Hypothesis 2.b: For the heterogeneous-income treatments, the spatial distribution of the different types will affect the transfer decisions – in the distribution where the high- and low-income type are alternated the transfers will be larger than where they are clustered. For testing these two hypothesis, we will compare the investment and transfer decisions in the treatments where the agents are distributed alternatively and clustered regarding their income type (treatment 3 vs treatment 4, and separately treatment 5 vs. treatment 6) Hypothesis 3.a: For the heterogeneous-income treatments, the common-chest transfer mechanism will give larger investments than the setting with direct transfers. Hypothesis 3.b: For the heterogeneous-income treatments, the common-chest transfer mechanism will give larger transfers from the high-income types to the low-incomes types than the direct transfer case. For testing these two hypothesis, we will compare the investment and transfer decisions in the treatments where the agents can make direct transfers and where the transfers are channelled through a common-chest mechanism (treatment 3 vs. treatment 5, and separately treatment 4 vs. treatment 6).
Secondary outcomes
Hypothesis 4: The overlapping-neighborhood setting triggers a spatial dynamic of provision levels in which indirect neighbors decisions of the previous period affect the investment decision in the current period. To gain insights into the dynamics of investment decisions across space, we compare the investment decisions in subsequent periods. Control variables from a short questionnaire at the end of the experiment: Demographics: Sex, age, field of study. Social preferences: risk-taking, prosociality. Task-related opinions: satisfaction with their own/their groupmates decisions, perceived generosity of their own/their groupmates decisions. Task comprehension: What do the participants think that was the subject of the study.

Cited by

No citation data available for this registration yet.

Keywords