Measuring Legal and Economic Controls

While transparency reports focus on platform moderation decisions, legal and economic controls operate through external mechanisms: government takedown requests, legal pressure, and economic restrictions. This activity explores how to measure these different forms of control and distinguish them from platform-initiated content moderation.

Background

The key distinction between platform controls and legal/economic controls lies in the source of the restriction:

• Platform controls reflect platforms’ own moderation decisions based on their terms of service, community guidelines, and business interests.
• Legal controls reflect external pressure from governments, legal systems, or third-party legal actors (copyright holders, courts, law enforcement).
• Economic controls restrict access or content through pricing, monetization, or infrastructure restrictions rather than direct blocking.

Understanding how to measure each type of control reveals different power structures and accountability mechanisms in online speech governance.

Activity 1: Lumen Database Analysis

The Lumen Database collects legal complaints about online content, including DMCA takedown notices, court orders, and government requests.

Task: Query the Lumen Database for DMCA notices related to a topic of interest (e.g., a specific website, a political issue, a type of content). Analyze the patterns you find:

• Who sends the most notices?
• What content is being targeted?
• Do the notices appear legitimate or potentially abusive?
• How quickly do platforms respond?
• Are there patterns by jurisdiction or time period?

Reflection: What does this reveal about how copyright law is used (or misused) to control online speech? How does this differ from platform moderation?

Activity 2: DSA Transparency Database Research

The DSA Transparency Database provides machine-readable data on content moderation decisions, including those made in response to government orders.

Task: Use the DSA Transparency Database to compare content moderation patterns across platforms:

• Compare removal rates between different platforms (e.g., YouTube vs TikTok)
• Examine what categories of content are most frequently removed
• Distinguish between platform-initiated removals and government orders
• Analyze whether decisions are automated or manual
• Track trends over time

Advanced: Apply for Research API access to conduct more sophisticated analysis.

Reflection: How do government-ordered removals differ from platform-initiated moderation? Which platforms are most transparent? What information is still missing?

Activity 3: Platform Policy Comparison

Task: Select three platforms (e.g., YouTube, Facebook/Meta, Twitter/X) and compare their transparency reports:

• Which platforms publish the most detailed data?
• Which countries or regions make the most government requests?
• How do compliance rates vary across platforms and countries?
• What categories of requests are most common?
• How has this changed over time?

Reflection: What explains the differences between platforms? Between countries? What might transparency reports not capture about government pressure?

Activity 4: Natural Experiment Analysis

Legal changes create opportunities to measure policy effects by comparing behavior before and after the change.

Task: Choose a recent legal or regulatory change:

• FOSTA/SESTA impact on platforms (2018)
• GDPR and website blocking of EU users (2018)
• Texas HB20 or Florida SB7072 (social media laws)
• DSA requirements (2024)
• State-level “age verification” laws

Research Questions:

• What data would you need to measure the law’s effects?
• What metrics would indicate impact?
• What confounding factors might complicate the analysis?
• How would you design a study to isolate the law’s effects?

Reflection: Can you find existing research or news reports about the law’s effects? Do they align with your expectations?

Activity 5: Economic Controls Investigation

Option A: Zero-Rating Study

Task: Research which services are zero-rated in different countries:

• Identify zero-rating programs in at least 3 countries
• What services are included? What’s excluded?
• What are the pricing differences?
• Can you find research on user behavior patterns?

Reflection: Does zero-rating appear to function primarily as data conservation or as a “walled garden” that limits information diversity?

Option B: Demonetization Analysis

Task: Monitor a sample of YouTube channels on controversial topics:

• Track which videos get demonetized
• Do certain topics correlate with higher demonetization rates?
• How do creators respond (content changes, platform migration)?
• What patterns emerge across different content categories?

Reflection: How does demonetization function as economic censorship? How does it differ from content removal?

Option C: App Availability Tracking

Task: Create a tool or use existing services to monitor app store availability:

• Select 10 apps (VPNs, messaging apps, news apps, circumvention tools)
• Check availability across 5+ countries
• Document any geographic restrictions
• Investigate reasons for unavailability

Reflection: How does infrastructure censorship through app stores differ from network-level blocking? Who has the power to restrict access?

Activity 6: Privacy Policy Evolution

Task: Use the Wayback Machine to track how a platform’s privacy policy or terms of service have changed over time:

• Select a major platform
• Compare versions from 2015, 2018, 2021, and 2024
• What has changed? What’s been added or removed?
• Do changes correlate with laws (GDPR, CCPA, DSA) or controversies?

Reflection: How do legal requirements shape platform policies? Do platforms go beyond legal minimums? How transparent are they about changes?

Reflection Questions

Think about the following questions:

• How do legal controls differ from platform controls? From technical controls?
• What are the challenges of measuring controls that don’t operate through direct blocking?
• Who has power over online speech, and how is that power exercised through different mechanisms?
• What information is missing from available data sources?
• Which forms of control are most visible? Most hidden?
• How do economic incentives shape what content is produced, distributed, and amplified online?