[Paper Review] Designing for Participation and Change in Digital Institutions

P. M. Krafft, Brian Keegan+1 more|arXiv (Cornell University)|Feb 23, 2019

Open Source Software Innovations58 references4 citations

TL;DR

This paper proposes integrating participatory constitutional layers—mechanisms enabling stakeholders to democratically amend institutional rules—into digital institutions like cryptocurrencies, cannabis informatics platforms, and Minecraft servers. Drawing on Elinor Ostrom’s resource governance framework, it demonstrates that such layers are essential for adaptive, ethical, and sustainable digital institutions, contrasting them with top-down, incentive-driven design models that exclude users from systemic change.

ABSTRACT

Whether we recognize it or not, the Internet is rife with exciting and original institutional forms that are transforming social organization on and offline. Issues of governance in these Internet platforms and other digital institutions have posed a challenge for software engineers, many of whom have little exposure to the relevant history or theory of institutional design. Here, we offer one useful framework with an aim to stimulate dialogue between computer scientists and political scientists. The dominant guiding practices for the design of digital institutions to date in human-computer interaction, computer-supported cooperative work, and the tech industry at large have been an incentive-focused behavioral engineering paradigm, a collection of atheoretical approaches such as A/B-testing, and incremental issue-driven software engineering. One institutional analysis framework that has been useful in the design of traditional institutions is the body of resource governance literature known as the "Ostrom Workshop". A key finding of this literature that has yet to be broadly incorporated in the design of many digital institutions is the importance of including participatory change process mechanisms in what is called a "constitutional layer" of institutional design---in other words, defining rules that allow and facilitate diverse stakeholder participation in the ongoing process of institutional design change. We explore to what extent this consideration is met or could be better met in three varied cases of digital institutions: cryptocurrencies, cannabis informatics, and amateur Minecraft server governance. Examining such highly varied cases allows us to demonstrate the broad relevance of constitutional layers in many different types of digital institutions.

Motivation & Objective

• To address the lack of inclusive, participatory mechanisms in the design of digital institutions, which often exclude users from systemic change.
• To challenge the dominant top-down, incentive-focused engineering paradigm in digital institution design that prioritizes behavioral nudges over democratic participation.
• To demonstrate the relevance and feasibility of Ostrom’s institutional design framework—particularly the constitutional layer—for diverse digital institutions.
• To show how participatory change mechanisms can enhance adaptability, fairness, and sustainability in digital systems.
• To argue that digital institutions must evolve from the bottom up through inclusive, rule-based amendment processes to avoid moral drift and user vulnerability.

Proposed method

• Adapting Ostrom’s institutional analysis framework, particularly the concept of a 'constitutional layer,' to digital institutions.
• Analyzing three distinct digital institutions: decentralized cryptocurrencies, cannabis informatics platforms, and amateur Minecraft server governance.
• Examining how stakeholder participation in rule amendment processes shapes institutional resilience and legitimacy.
• Comparing the effectiveness of 'vote with your feet' exit mechanisms in Minecraft versus the immutability and high barriers to exit in cryptocurrencies.
• Using qualitative case study analysis to assess the presence, design, and impact of participatory change mechanisms in each institution.
• Contrasting top-down design models (e.g., Kraut & Resnick’s community engineering) with Ostrom-inspired, participatory models to highlight structural differences in governance capacity.

Experimental results

Research questions

• RQ1 To what extent do existing digital institutions incorporate mechanisms that allow stakeholders to participate in rule amendment processes?
• RQ2 How do participatory constitutional layers affect the sustainability, adaptability, and legitimacy of digital institutions?
• RQ3 Why do exit mechanisms like 'vote with your feet' succeed in some digital institutions (e.g., Minecraft) but fail in others (e.g., cryptocurrencies)?
• RQ4 What are the structural and technical barriers that prevent the integration of participatory change processes in digital institutions designed through top-down engineering?
• RQ5 How can Ostrom’s resource governance framework be meaningfully translated into the design of digital institutions with code-based enforcement?

Key findings

• Digital institutions such as cryptocurrencies, cannabis informatics platforms, and Minecraft servers often lack formal constitutional layers that enable stakeholder-driven rule changes.
• In Minecraft server governance, successful administrators learn to implement participatory policy-making and modular code systems on the job, demonstrating bottom-up institutional adaptation.
• The 'vote with your feet' mechanism functions effectively in Minecraft due to low exit costs and high competition among servers, enabling ecosystem-wide change through user mobility.
• In contrast, the immutability and high transaction costs of cryptocurrencies severely limit user exit and rule amendment, making participatory change nearly impossible under current designs.
• Top-down engineering models, such as those in Building Successful Online Communities, prioritize behavioral optimization and user retention over democratic participation, leading to institutions that are inaccessible and unaccountable to users.
• The absence of constitutional layers renders digital institutions vulnerable to moral drift, power imbalances, and systemic abuse, especially when designers and operators are a small, homogeneous group.