Why Scientology, Jonestown and the FLDS All Score in the High-30s — and Why That's a Limit, Not an Equivalence
The CLCI maxes out at 40. That ceiling forces qualitatively different harms — financial extraction, mass-casualty violence, systematic child abuse — into the same numeric band. Here is how to read the 31–40 entries without confusing the score with the lived consequence.
When you sort the CLCI Hub dataset by total score and look at the top of the list, you find a problem the formula cannot solve.
The Church of Scientology scores 37/40. The People's Temple, which murdered 918 of its own members at Jonestown in 1978, scores 40/40. Heaven's Gate, whose 39 members died by group suicide in 1997, scores 40. Aum Shinrikyo, which released sarin gas on the Tokyo subway in 1995 and killed 13 people, scores 40. The historical Russian Skoptsy, who required surgical castration of male adherents and breast or genital mutilation of female ones, score 38.
If you read those scores as a granular ranking — Scientology is "8% less harmful" than the People's Temple — you have misread them. They are not granular at the top. They cannot be.
Why the ceiling is real
The CLCI is built on Steven Hassan's BITE model: four axes (Behavior, Information, Thought, Emotional) scored from 0 to 10 each, plus a small signed modifier from −5 to +5. Total maximum: 40. Total minimum: 0.
That formula can capture a great deal of structural variation in the moderate ranges. A group scored 14/40 looks meaningfully different from one scored 22/40. A 22 looks meaningfully different from a 28. The formula does work it is designed to do, in the band it was designed to do it in.
What the formula was not designed to do is differentiate qualitatively different categories of extreme harm from each other. Once a group reaches the maximum on each BITE axis, the only remaining lever is the modifier — and the modifier is capped at +3 in practice. There is nowhere left to put "918 deaths" relative to "decades-long pattern of forced confessional extraction".
What this means for reading the dataset
Three rules of thumb:
- Treat the 31–40 band as a category, not a ranking. The category means "this group exhibits destructive control across multiple BITE axes, with documented major harm." The relative position of two entries inside that band is not load-bearing.
- Read the body and timeline, not just the badge. The qualitative differentiation that the number cannot express is in the prose. The Scientology body documents disconnection, financial extraction, and the Suppressive Person designation. The People's Temple body documents the events of November 18 1978. Both are visible. The score badge tells you neither.
- Don't compare across categories of harm without thinking about what the modifier captures. A group that scored +3 on the modifier for "documented multi-victim founder sexual abuse" and a group that scored +2 for "mass-casualty violence" are not commensurable on a single axis. They are scored as the harms that they are.
What we considered, and why we didn't change the formula
We thought about extending the modifier range to ±10 to give the top of the scale more headroom. The reason we didn't is that the BITE model's design integrity is more valuable than fine-grained numeric differentiation at the top. Steven Hassan's framework is widely used because the four 0–10 axes plus a small modifier are easy to understand, easy to score, and produce defensible inter-rater agreement. A custom ±10 modifier specific to this site would solve a presentation problem at the cost of methodological coherence.
We also considered adding a separate harmTier field — fatalities | physical-abuse | psychological | financial — independent of the CLCI number. We may still do this. The arguments against are that any such field would be subjective at the boundaries and would invite the same comparison-across-categories confusion the score already causes. The arguments for are that several entries genuinely do warrant a "documented mass-casualty event" indicator that the body text alone communicates poorly to a casual reader.
For now: the 31–40 band has a known limit, and we have documented it on the /about page. When you read a high-30s score, read the body.
This is educational, not legal or clinical advice. If you need support, consult a licensed therapist or contact ICSA.