This study examines behavior in a voluntary resource sharing game that incorporates endogenous network formation, an increasingly common incentive problem in contemporary digital economies. The experiment was conducted in two parts with 184 subjects across 46 groups. In the first part, each subject was assigned a unique ID and played the game for 15 rounds without any information about the individual decisions made by others in their group. In the second part, subjects were randomly reassigned IDs and played another 15 rounds with three baseline sessions and five treatment sessions. In the treatment sessions, subjects were shown how much benefit they received from each other subject in their group in the previous round; this information facilitated direct reciprocity which was not present in baseline sessions. After completing both parts of the experiment, subjects completed questionnaires to elicit behavioral characteristics. The study uses evidence from this laboratory experiment to explore information design for this environment and compare maximum likelihood estimation (MLE) and modified partial likelihood estimation (MPLE) panel estimators. The results show that providing subjects with information about past behavior has a significant impact on efficiency and performance in the market. The study estimates the effects of the treatment on several outcome measures using structural methods to investigate underlying behavioral traits that generate these observed treatment effects. Reciprocity is decomposed into direct reciprocity and generalized or altruistic reciprocity; it is found that players react to new information by focusing on reciprocity which helps them coordinate to reach mutually beneficial outcomes. Direct and indirect reciprocity are not perfect substitutes either, further helping players support efficient outcomes. Additionally, network configurations that emerge from data are stable under parameter estimates while individual heterogeneity is another benefit of repeated-choice modeling paradigm explored by this study. Finally, robustness analysis is performed on estimators' performance on these data to provide insights into how information design can affect behavior in environments with endogenous network formation as well as highlight important implications for digital economies.
- - The study examines behavior in a voluntary resource sharing game with endogenous network formation
- - 184 subjects across 46 groups participated in the experiment conducted in two parts
- - In the second part, subjects were shown how much benefit they received from each other subject in their group in the previous round to facilitate direct reciprocity
- - Providing information about past behavior has a significant impact on efficiency and performance in the market
- - Reciprocity is decomposed into direct and generalized/altruistic reciprocity, which helps players coordinate for mutually beneficial outcomes
- - Network configurations that emerge from data are stable under parameter estimates while individual heterogeneity is another benefit of repeated-choice modeling paradigm explored by this study
- - Robustness analysis is performed on estimators' performance on these data to provide insights into how information design can affect behavior in environments with endogenous network formation as well as highlight important implications for digital economies.
Summary: The study looked at how people share resources in a game where they can choose who to share with. 184 people played the game in groups. In the second part of the game, players were told how much benefit they got from others in their group before. This helped them work together better. Sharing information about past behavior made the game more efficient and everyone did better. The study found that when people work together, they can get good results.
Definitions:
- Behavior: How someone acts or behaves
- Voluntary: When you choose to do something on your own
- Reciprocity: When two people help each other out
- Efficiency: Doing things well and quickly
- Generalized/Altruistic reciprocity: Helping others without expecting anything in return
- Network formation: How people connect with each other
- Robustness analysis: Checking if something works well even when there are changes or problems
Exploring Information Design and Behavioral Traits in Voluntary Resource Sharing Games with Endogenous Network Formation
In the digital economy, incentive problems are becoming increasingly common. This study examines behavior in a voluntary resource sharing game that incorporates endogenous network formation to explore information design and compare maximum likelihood estimation (MLE) and modified partial likelihood estimation (MPLE) panel estimators. The experiment was conducted over two parts with 184 subjects across 46 groups. In the first part, each subject was assigned a unique ID and played the game for 15 rounds without any information about the individual decisions made by others in their group. In the second part, subjects were randomly reassigned IDs and played another 15 rounds with three baseline sessions and five treatment sessions.
Information Design
In the treatment sessions, subjects were shown how much benefit they received from each other subject in their group in the previous round; this information facilitated direct reciprocity which was not present in baseline sessions. After completing both parts of the experiment, subjects completed questionnaires to elicit behavioral characteristics. The study uses evidence from this laboratory experiment to investigate underlying behavioral traits that generate observed treatment effects on several outcome measures using structural methods.
Reciprocity
The results show that providing subjects with information about past behavior has a significant impact on efficiency and performance in the market. Reciprocity is decomposed into direct reciprocity and generalized or altruistic reciprocity; it is found that players react to new information by focusing on reciprocity which helps them coordinate to reach mutually beneficial outcomes. Direct and indirect reciprocity are not perfect substitutes either, further helping players support efficient outcomes. Additionally, network configurations that emerge from data are stable under parameter estimates while individual heterogeneity is another benefit of repeated-choice modeling paradigm explored by this study.
Robustness Analysis
Finally, robustness analysis is performed on estimators' performance on these data to provide insights into how information design can affect behavior in environments with endogenous network formation as well as highlight important implications for digital economies moving forward. This research paper provides valuable insight into understanding incentives within digital economies through exploring behavior within voluntary resource sharing games incorporating endogenous network formation as well as comparing MLEs versus MPLEs panel estimators when analyzing data collected from such experiments..