Exercises: Designing Data Futures

Claude (AI-Generated Textbook)

Exercises: Designing Data Futures

These exercises progress from concept checks to challenging applications, including Python programming exercises. Estimated completion time: 4-5 hours.

Difficulty Guide: - ⭐ Foundational (5-10 min each) - ⭐⭐ Intermediate (10-20 min each) - ⭐⭐⭐ Challenging (20-40 min each) - ⭐⭐⭐⭐ Advanced/Research (40+ min each)

Part A: Conceptual Understanding ⭐

Test your grasp of core concepts from Chapter 39.

A.1. Define the "participation deficit" in data governance. Why does the chapter argue that the participation deficit is a specific instance of the power asymmetry theme?

A.2. Compare and contrast data cooperatives, data trusts, and data commons. What is the key structural difference in governance authority among the three models? Under what conditions is each model most appropriate?

A.3. Explain what a citizen assembly is and how it differs from other forms of public consultation (e.g., public comment periods, advisory committees, opinion surveys). What makes citizen assemblies particularly well-suited to data governance decisions?

A.4. Define speculative design. How does it differ from prediction or forecasting? Why does the chapter argue that "every governance framework is also a kind of fiction"?

A.5. What is "prefigurative governance"? Explain the relationship between prefigurative governance and institutional change. Are they substitutes or complements?

A.6. In the GovernanceSimulator, the Gini coefficient measures inequality in benefit distribution. Explain what a Gini coefficient of 0 means, what a Gini coefficient of 1 means, and why this metric is useful for comparing governance models.

A.7. The chapter describes "hope as a political practice." What does this mean? How does it differ from optimism, and why does the chapter argue that hope is a necessary condition for effective governance?

Part B: Applied Analysis ⭐⭐

Analyze scenarios, arguments, and real-world situations.

B.1. A city government proposes creating a citizen assembly to advise on the governance of its smart city infrastructure. The assembly will consist of 50 randomly selected residents who meet for four weekends to hear expert testimony and deliberate. City officials say the assembly's recommendations will be "carefully considered." Civil society organizations argue that advisory-only status makes the assembly performative. Evaluate both positions. Under what conditions would you consider this assembly genuinely participatory?

B.2. A group of gig workers proposes forming a data cooperative to collectively govern the data generated by their work on ride-share and delivery platforms. The cooperative would negotiate with platforms for access to trip data, pay data, and algorithmic decision data. Analyze this proposal using Elinor Ostrom's eight principles for commons governance. Which principles will be most challenging to implement, and why?

B.3. Section 39.4 describes speculative design methods, including "design fiction" and "backcasting." Choose one data governance challenge from a previous chapter (e.g., algorithmic bias, consent mechanisms, surveillance) and write a brief design fiction (200-300 words) that depicts how this challenge has been resolved in 2045. Then identify three specific governance decisions that would need to happen between now and 2045 for your fiction to become reality.

B.4. Eli's Community Data Governance Charter (Section 39.7) includes provisions for community consent, data councils, sensor-free zones, and sunset clauses. A critic argues: "This is a nice idea, but a neighborhood charter has no legal force. The city can ignore it." Evaluate this critique. What strategies could Eli's community use to give the charter practical power beyond moral authority?

B.5. Mira's Reformed Governance Framework for VitraMed was adopted by the board as a two-year pilot, but the executive compensation provision was deferred. Analyze this outcome. Why might the board accept the five pillars but defer compensation? What does this reveal about the limits of governance reform within corporate structures?

B.6. The chapter argues that participatory governance produces better governance than expert-designed governance. Present the strongest counter-argument to this claim. Under what conditions might expert-designed governance be superior to participatory governance? Are there trade-offs between participation and effectiveness?

Part C: Python Exercises ⭐⭐-⭐⭐⭐

These exercises require working with the GovernanceSimulator from Section 39.5.

C.1. ⭐⭐ Run the Simulation. Using the GovernanceSimulator class and create_example_community() function from Section 39.5, run the simulation and examine the output. Answer the following:

(a) Which governance model produces the highest total benefit? Which produces the lowest Gini coefficient?
(b) Which stakeholder benefits most under the Corporate Centralized model? Which benefits least? Why?
(c) Which model results in the fewest stakeholders below their benefit threshold?
(d) Examine the privacy protection scores. Which model provides the most equal privacy protection? Which provides the highest average privacy protection?

C.2. ⭐⭐ Custom Community. Create a new community of at least five stakeholders representing a different scenario — for example, a rural farming community, a university campus, or a hospital system. Assign attribute values (data_contribution, technical_literacy, political_influence, privacy_preference, benefit_threshold) that reflect the real-world dynamics of your chosen scenario. Run the simulation and write a one-paragraph interpretation of the results.

# Starter code
my_community = [
    Stakeholder("Stakeholder 1", "group", dc, tl, pi, pp, bt),
    # ... add at least 4 more
]
sim = GovernanceSimulator(my_community, seed=42)
results = sim.run_all_models()
sim.display_comparison(results)

C.3. ⭐⭐⭐ Modify the Cooperative Model. The cooperative_democratic method applies a coordination cost of 0.85 (15% efficiency loss). Modify the method to accept a coordination_cost parameter. Then run the simulation with coordination costs of 0.75, 0.85, 0.90, and 0.95. At what point does the cooperative model's total benefit exceed the regulatory model's? What does this tell you about the relationship between coordination efficiency and governance model viability?

# Hint: modify the method signature and replace the hardcoded value
def cooperative_democratic(self, coordination_cost: float = 0.85) -> GovernanceOutcome:
    # ... modify the method body accordingly

C.4. ⭐⭐⭐ Add a New Governance Model. Design and implement a fifth governance model: corporate_regulated — a model where corporate governance is constrained by regulatory minimums. Benefits should be distributed like the corporate model but with a regulatory floor; privacy should have a minimum threshold; and voice should combine corporate influence with regulatory access rights.

def corporate_regulated(self) -> GovernanceOutcome:
    """Simulate a corporate governance model with regulatory constraints."""
    # Your implementation here
    pass

Run the simulation with all five models and compare the results. How does the hybrid corporate-regulatory model compare to the pure models?

C.5. ⭐⭐⭐ Sensitivity Analysis. Write a function that systematically varies one stakeholder's political_influence from 10 to 90 (in steps of 10) while holding all other attributes constant, and records the stakeholder's benefit under each governance model for each value. Plot or tabulate the results. Which governance model is most sensitive to political influence? Which is least sensitive?

def sensitivity_analysis(community, stakeholder_index, attribute, values, seed=42):
    """Vary one attribute and track outcomes across governance models."""
    results = {}
    for val in values:
        # Create a modified community with the varied attribute
        # Run simulation and record outcomes
        pass
    return results

C.6. ⭐⭐⭐⭐ Dynamic Simulation. Extend the GovernanceSimulator to support multiple rounds of governance. After each round, stakeholders adjust their behavior: those below their benefit threshold increase their political activity (political_influence increases by 5), and those significantly above their threshold become less active (political_influence decreases by 3). Run the simulation for 10 rounds under each governance model. How do the outcomes evolve over time? Which models produce stable distributions, and which produce increasing inequality or instability?

Part D: Synthesis & Critical Thinking ⭐⭐⭐

These questions require you to integrate multiple concepts from Chapter 39 and earlier chapters.

D.1. The GovernanceSimulator demonstrates that governance structure shapes outcomes — the same community receives different benefits under different models. But the simulation's results depend on its assumptions (weight parameters, coordination costs, noise levels). Does this dependence on assumptions undermine the simulation's usefulness, or does the act of debating assumptions have governance value in itself? Write a 300-word response.

D.2. Connect the participatory governance models from this chapter to the Global South perspectives from Chapter 37. Could data cooperatives and citizen assemblies address the challenges of data colonialism? Under what conditions? What obstacles would participatory governance face in contexts marked by infrastructure dependency and limited institutional capacity?

D.3. Dr. Adeyemi asks: "If the systems that govern data are designed by people, who should those people be?" Write a response (400-500 words) that draws on the participatory models from this chapter, the democratic oversight challenges from Chapter 36, and the ethical frameworks from Chapter 6.

D.4. Evaluate the chapter's closing argument that "hope is a political practice." Is this a rigorous intellectual claim, or is it motivational rhetoric? What evidence from the chapter's case studies (Barcelona, Aotearoa New Zealand) supports or challenges the claim? Write a 300-word assessment.

Part E: Research & Extension ⭐⭐⭐⭐

Open-ended projects for students seeking deeper engagement.

E.1. Data Cooperative Case Study. Research one existing data cooperative (e.g., MIDATA in Switzerland, Driver's Seat Cooperative in the US, Salus Coop in Barcelona). Write a 1,000-word case study covering: (a) the cooperative's structure and governance, (b) the data it manages, (c) the benefits it has delivered to members, (d) the challenges it has faced, and (e) the lessons it offers for data governance reform.

E.2. Citizen Assembly Design. Design a citizen assembly for a specific data governance question in your community (e.g., "Should our city use facial recognition in public spaces?" or "How should student data be governed by our university?"). Your design should specify: (a) how assembly members are selected, (b) what information and expert testimony they receive, (c) the deliberation structure, (d) how recommendations are formulated, and (e) the relationship between the assembly's recommendations and decision-making authority. Write a 1,200-word proposal.

E.3. Speculative Design Portfolio. Create three design fiction artifacts from a speculative data governance future (e.g., a privacy policy from a neural data cooperative in 2040, a news article about the first conviction under a cognitive liberty law, an advertisement for a community-governed digital twin). Each artifact should be 300-500 words and include a brief critical reflection (200 words) on the governance assumptions embedded in the fiction.

Solutions

Selected solutions are available in appendices/answers-to-selected.md.