The LEWS System

A comprehensive look at the Lock-in Early Warning System architecture and functionality

Understanding Lock-in in Animal Agriculture

Lock-in occurs when technology, infrastructure, regulation, and norms converge so the system becomes self-reinforcing and extremely hard to reverse. The LEWS system is designed to identify the early warning signs of this process before it becomes irreversible.

Historical Context

One of the most prominent historical examples of lock-in in animal agriculture is the adoption of battery cages between 1950–1970. Once the system became entrenched, it proved extremely difficult to reverse despite growing awareness of animal welfare issues.

The Seven Core Variables

LEWS tracks seven key variables that indicate the likelihood of lock-in:

Population Scale

Based on data from Our World in Data, FAO Aquaculture Databases, and Rethink Priorities. Tracks annual slaughter numbers and standing populations.

Sentience Probability

Derived from Rethink Priorities Sentience Project, Birch/Browning/Crump papers, and EFSA evaluations. Expressed as 0-100% probability.

Suffering Intensity

Based on Welfare Footprint Project data and peer-reviewed ethology. Converted to a -10 to +10 valence scale for usability.

Industry Momentum

Quantified through investment data, corporate announcements, FAO growth rates, and patent activity. Converted to qualitative scores 0-100.

Advocacy Gap

Based on Animal Charity Evaluators, Open Philanthropy, and a directory of organizations working on each species. Shows how much advocacy exists.

Lock-in Signals

Derived from political science theory: regulatory capture, harmonization, capital fixity, path dependence, and standardization.

Uncertainty

Explicit uncertainty ranges that reflect model uncertainty, epistemic uncertainty, behavioral uncertainty, and biological uncertainty.

How LEWS Works

The LEWS system uses a linear weighted model to calculate a 0-100 lock-in risk score:

risk_score = 0.30 × scale_norm + 0.20 × suffering_norm + 0.15 × sentience_norm + 0.20 × momentum_norm + 0.10 × advocacy_gap_norm + 0.05 × time_to_lockin_norm

Each variable is normalized to 0-1 and weighted based on empirical relevance. The system then applies uncertainty ranges to provide confidence intervals.

Output Components

  • Risk Range: A confidence interval like "Lock-In: 72 (58–86)"
  • Stage Classification: Research → Scaling → Lock-in phases
  • Intervention Urgency: Monitor / Act Soon / Act Now recommendations
  • Key Metrics: Suffering hours, animal numbers, and other relevant statistics

Technology Comparison

LEWS allows for comparison between different animal farming technologies. For example, it might determine that "Current shrimp farming ≈ chicken farming in 1950", providing historical context for current risk assessment.

Data Sources

LEWS is grounded in authoritative data sources including:

  • Our World in Data for slaughter numbers
  • FAO Aquaculture Databases
  • Rethink Priorities (insect, shrimp, fish estimations)
  • Rowe (2020–2023) for insect-scale estimates
  • Mood & Brooke for fish and shrimp data
  • Welfare Footprint Project for suffering intensity
  • Investment databases like Crunchbase and AgFunder