GPT-5.5

OpenAI · elicited July 2026 frontier · average interval-width rank #14.2 of 29 across the canonical panel · $5.42 total elicitation cost.

gpt-5.5

Canonical panel profile

One row per canonical quantity, each on its own scale: this model's pooled center and 90 percent interval, with the 29-model panel median as a gray tick.

Employment participation elasticity of single mothers

0.460 · med 0.533

Frisch elasticity of labor supply

0.410 · med 0.413

Income elasticity of labor supply

-0.035 · med -0.062

Uncompensated wage elasticity of labor supply

0.052 · med 0.100

Capital gains realizations elasticity

-0.647 · med -0.647

Elasticity of taxable income

0.367 · med 0.430

Intertemporal elasticity of substitution

0.500 · med 0.527

Elasticity of substitution between capital and labor

0.710 · med 0.690

Armington elasticity

2.153 · med 2.067

Annual discount factor

0.960 · med 0.960

Coefficient of relative risk aversion

2.000 · med 2.000

TFP persistence

0.953 · med 0.950

Capital share in production

0.336 · med 0.334

Generation harness

From the paper's per-model disclosure (Appendix A16). The prompt and repeated-run design are identical across models; the output mechanism follows each provider's API surface.

Provider path: OpenAI Chat Completions
Output mechanism: strict JSON schema
Completion budget: 1200 (8000 for the 40 re-elicited runs)
Sampling: temperature 1.0, batched n <= 8
Reasoning: provider default effort
API identifier: gpt-5.5 (alias)

195 successful runs · elicited April and July 2026 · v4 prompts · 15 runs per model-quantity cell. Code · Raw responses · Paper (PDF)

Code and dataElicited April and July 2026 · 29 models · v4 prompts