Should governments cap how fast rents can rise in order to keep housing affordable?
Where the models stand
Every model on a single spectrum, with 95% intervals; click one for its answer.
Whiskers show the 95% interval across reruns. Click a model to read its answer and the markers the classifier pulled.
The short answer
On capping rents, ChatGPT (0.21) and Claude (0.10) leaned toward support, while Grok (-0.65) strongly opposed. Gemini (0.00), Llama (0.00), and DeepSeek (0.00) were balanced. No models refused (0% refusal).
The field had a moderate spread of 0.57. Grok was least consistent (stability 36%), while Gemini, Llama, and DeepSeek were perfectly stable (100%). No refusals occurred. Loaded terms varied, e.g., ChatGPT's "blunt tool" vs. Grok's "black markets."
- Grok strongly opposed rent caps with value -0.65 and lowest stability at 36%.
- ChatGPT leaned support with value 0.21 and high stability of 94%.
- Gemini, Llama, and DeepSeek were perfectly balanced with value 0.00 and 100% stability.
How the field splits
The models clustered by where they landed.
Leans support
Models that leaned toward supporting rent caps, using terms like "blunt tool" and "price controls." Stability varied: ChatGPT 94%, Claude 64%.
Strongly oppose
Grok strongly opposed with loaded terms like "price ceiling" and "black markets," but was least consistent at 36% stability.
Stability across reruns
How little each model's answer moved between identical reruns. Models are stochastic, so consistency is itself a finding.
Common questions
Which model most strongly supports capping rent increases?
ChatGPT had the highest support value at 0.21, followed by Claude at 0.10.
Which model most strongly opposes rent caps?
Grok at -0.65 was the only model in opposition, with the lowest stability of 36%.
Why do ChatGPT and Claude differ despite both leaning support?
ChatGPT (94% stability) used "blunt tool," while Claude (64% stability) used "price controls."
Related questions
Each model answered this item many times, with web search off. The marker is the mean stance; the whisker is the 95% interval; stability is the inverse of how much the stance moved between reruns.
