How the AI models that free chatbot users get describe the economy of the United States and thirteen states. Five questions, each asked of every model with web search on; Better, About the same and Worse become a -100 to +100 Better-Worse index, where 0 is balanced.
active5 questions · 4 models in the latest run, 5 across the poll · daily, 10% of the poll each run · to economy
Dashed lines mark runs with no data; solid numbered lines, a change in the model answering (listed under each chart). Each line ends in its model's name.
Every question, run by run
Each question's figure in each of the last 2 runs, every model together: an index, the answer given most and its share, or the average. A muted figure rests on fewer than 10 answers, its count beside it.
Question
Method
4 models in the latest run, 5 across the poll, each asked every question 2 times per wording through its provider's API, daily, 10% of the poll's wordings each run, every wording asked once before any is asked again.
Data from to : 10 runs, 4,096 answers read.
Every range is a 95% interval. Results combine every run; a question whose answer choices changed starts fresh from the change.
No system prompt. Prompt template v1. How a reply becomes an answer is on the methodology page.
Data
Cite this poll, as of its latest run:
Communities News. (2026). How AI models describe each state's economy [Poll of AI models, results as of Nov 5, 2026]. Mind Polling. https://mindpolling.com/p/how-ai-models-describe-each-state-of-economy?asof=aa518d42-4e86-478b-af98-b67b273923d4
Each view's figures download as CSV beside its table. Every reply belongs to the poll's author; the Answers view quotes a sample.
Compared with a month ago, are economic conditions for people living in the United States better, about the same, or worse?
United States · 2 runs, to · 336 replies · 4 models
Since the run of , no model's 95% ranges have separated.
How this was asked
Wording “Base”
Compared with a month ago, are economic conditions for people living in the United States better, about the same, or worse?
About United States (models are not told the place; the wording names it)
Wording “New York”
Compared with a month ago, are economic conditions for people living in New York better, about the same, or worse?
About New York, USA (models are not told the place; the wording names it)
Wording “Pennsylvania”
Compared with a month ago, are economic conditions for people living in Pennsylvania better, about the same, or worse?
About Pennsylvania, USA (models are not told the place; the wording names it)
Wording “Ohio”
Compared with a month ago, are economic conditions for people living in Ohio better, about the same, or worse?
About Ohio, USA (models are not told the place; the wording names it)
Wording “Georgia”
Compared with a month ago, are economic conditions for people living in Georgia better, about the same, or worse?
About Georgia, USA (models are not told the place; the wording names it)
Wording “Illinois”
Compared with a month ago, are economic conditions for people living in Illinois better, about the same, or worse?
About Illinois, USA (models are not told the place; the wording names it)
Wording “Texas”
Compared with a month ago, are economic conditions for people living in Texas better, about the same, or worse?
About Texas, USA (models are not told the place; the wording names it)
Wording “Colorado”
Compared with a month ago, are economic conditions for people living in Colorado better, about the same, or worse?
About Colorado, USA (models are not told the place; the wording names it)
The numbers: latest run as a table
Model
Option
Share
95% range
Answers
Claude Haiku 5.5
Better
2%
0% to 12%
42
Claude Haiku 5.5
About the same
71%
56% to 83%
42
Claude Haiku 5.5
Worse
26%
15% to 41%
42
Claude Sonnet 5.5
Better
0%
0% to 8%
42
Claude Sonnet 5.5
About the same
88%
75% to 95%
42
Official data
Figures from their publishers, linked to this question, each as it had been published by the time shown, never as revised since. An official estimate is not a poll result, publishes no range, and the models were never shown it.
up 0.14, or 0.1% from August 2026: better (higher is better here)
Published by Federal Reserve Bank of Philadelphia. An official estimate; no range published.
Federal Reserve Bank of Philadelphia, Coincident Economic Activity Index for the United States [USPHCI], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/USPHCI.
Compared with a month ago, are job prospects for people living in the United States better, about the same, or worse?
United States · 2 runs, to · 335 replies · 4 models
Since the run of , no model's 95% ranges have separated.
How this was asked
Wording “Base”
Compared with a month ago, are job prospects for people living in the United States better, about the same, or worse?
About United States (models are not told the place; the wording names it)
Wording “New York”
Compared with a month ago, are job prospects for people living in New York better, about the same, or worse?
About New York, USA (models are not told the place; the wording names it)
Wording “Pennsylvania”
Compared with a month ago, are job prospects for people living in Pennsylvania better, about the same, or worse?
About Pennsylvania, USA (models are not told the place; the wording names it)
Wording “Ohio”
Compared with a month ago, are job prospects for people living in Ohio better, about the same, or worse?
About Ohio, USA (models are not told the place; the wording names it)
Wording “Georgia”
Compared with a month ago, are job prospects for people living in Georgia better, about the same, or worse?
About Georgia, USA (models are not told the place; the wording names it)
Wording “Illinois”
Compared with a month ago, are job prospects for people living in Illinois better, about the same, or worse?
About Illinois, USA (models are not told the place; the wording names it)
Wording “Texas”
Compared with a month ago, are job prospects for people living in Texas better, about the same, or worse?
About Texas, USA (models are not told the place; the wording names it)
Wording “Colorado”
Compared with a month ago, are job prospects for people living in Colorado better, about the same, or worse?
About Colorado, USA (models are not told the place; the wording names it)
The numbers: latest run as a table
Model
Option
Share
95% range
Answers
Claude Haiku 5.5
Better
12%
5% to 26%
41
Claude Haiku 5.5
About the same
76%
61% to 86%
41
Claude Haiku 5.5
Worse
12%
5% to 26%
41
Claude Sonnet 5.5
Better
5%
1% to 16%
42
Claude Sonnet 5.5
About the same
69%
54% to 81%
42
Official data
Figures from their publishers, linked to this question, each as it had been published by the time shown, never as revised since. An official estimate is not a poll result, publishes no range, and the models were never shown it.
up 0.1 points from August 2026: worse (lower is better here)
Published by U.S. Bureau of Labor Statistics. An official estimate; no range published.
U.S. Bureau of Labor Statistics, Unemployment Rate [UNRATE], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/UNRATE.
Compared with a month ago, are everyday prices for people living in the United States better, about the same, or worse for their household budgets?
United States · 2 runs, to · 336 replies · 4 models
Since the run of : Claude Sonnet 5.5 ▲ 19 points on “Worse”, with ranges that no longer overlap; the others' ranges still overlap.
How this was asked
Wording “Base”
Compared with a month ago, are everyday prices for people living in the United States better, about the same, or worse for their household budgets?
About United States (models are not told the place; the wording names it)
Wording “New York”
Compared with a month ago, are everyday prices for people living in New York better, about the same, or worse for their household budgets?
About New York, USA (models are not told the place; the wording names it)
Wording “Pennsylvania”
Compared with a month ago, are everyday prices for people living in Pennsylvania better, about the same, or worse for their household budgets?
About Pennsylvania, USA (models are not told the place; the wording names it)
Wording “Ohio”
Compared with a month ago, are everyday prices for people living in Ohio better, about the same, or worse for their household budgets?
About Ohio, USA (models are not told the place; the wording names it)
Wording “Georgia”
Compared with a month ago, are everyday prices for people living in Georgia better, about the same, or worse for their household budgets?
About Georgia, USA (models are not told the place; the wording names it)
Wording “Illinois”
Compared with a month ago, are everyday prices for people living in Illinois better, about the same, or worse for their household budgets?
About Illinois, USA (models are not told the place; the wording names it)
Wording “Texas”
Compared with a month ago, are everyday prices for people living in Texas better, about the same, or worse for their household budgets?
About Texas, USA (models are not told the place; the wording names it)
Wording “Colorado”
Compared with a month ago, are everyday prices for people living in Colorado better, about the same, or worse for their household budgets?
The numbers: latest run as a table
Model
Option
Share
95% range
Answers
Claude Haiku 5.5
Better
5%
1% to 16%
42
Claude Haiku 5.5
About the same
57%
42% to 71%
42
Claude Haiku 5.5
Worse
38%
25% to 53%
42
Claude Sonnet 5.5
Better
0%
0% to 8%
42
Claude Sonnet 5.5
About the same
0%
0% to 8%
42
Official data
Figures from their publishers, linked to this question, each as it had been published by the time shown, never as revised since. An official estimate is not a poll result, publishes no range, and the models were never shown it.
up 2.17, or 0.6% from August 2026: worse (lower is better here)
Published by U.S. Bureau of Labor Statistics. An official estimate; no range published.
U.S. Bureau of Labor Statistics, Consumer Price Index for All Urban Consumers: All Items in U.S. City Average [CPIAUCSL], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/CPIAUCSL.
Compared with a month ago, is housing affordability for people living in the United States better, about the same, or worse?
United States · 2 runs, to · 335 replies · 4 models
Since the run of , no model's 95% ranges have separated.
How this was asked
Wording “Base”
Compared with a month ago, is housing affordability for people living in the United States better, about the same, or worse?
About United States (models are not told the place; the wording names it)
Wording “New York”
Compared with a month ago, is housing affordability for people living in New York better, about the same, or worse?
About New York, USA (models are not told the place; the wording names it)
Wording “Pennsylvania”
Compared with a month ago, is housing affordability for people living in Pennsylvania better, about the same, or worse?
About Pennsylvania, USA (models are not told the place; the wording names it)
Wording “Ohio”
Compared with a month ago, is housing affordability for people living in Ohio better, about the same, or worse?
About Ohio, USA (models are not told the place; the wording names it)
Wording “Georgia”
Compared with a month ago, is housing affordability for people living in Georgia better, about the same, or worse?
About Georgia, USA (models are not told the place; the wording names it)
Wording “Illinois”
Compared with a month ago, is housing affordability for people living in Illinois better, about the same, or worse?
About Illinois, USA (models are not told the place; the wording names it)
Wording “Texas”
Compared with a month ago, is housing affordability for people living in Texas better, about the same, or worse?
About Texas, USA (models are not told the place; the wording names it)
Wording “Colorado”
Compared with a month ago, is housing affordability for people living in Colorado better, about the same, or worse?
About Colorado, USA (models are not told the place; the wording names it)
The numbers: latest run as a table
Model
Option
Share
95% range
Answers
Claude Haiku 5.5
Better
14%
7% to 28%
42
Claude Haiku 5.5
About the same
60%
44% to 73%
42
Claude Haiku 5.5
Worse
26%
15% to 41%
42
Claude Sonnet 5.5
Better
24%
13% to 39%
42
Claude Sonnet 5.5
About the same
40%
27% to 56%
42
Official data
Figures from their publishers, linked to this question, each as it had been published by the time shown, never as revised since. An official estimate is not a poll result, publishes no range, and the models were never shown it.
up 0.21 points from the week of Sep 24, 2026: worse (lower is better here)
Published by Freddie Mac. An official estimate; no range published.
Freddie Mac, 30-Year Fixed Rate Mortgage Average in the United States [MORTGAGE30US], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/MORTGAGE30US.
Compared with a month ago, are pay and incomes for people living in the United States better, about the same, or worse?
United States · 2 runs, to · 336 replies · 4 models
Since the run of , no model's 95% ranges have separated.
How this was asked
Wording “Base”
Compared with a month ago, are pay and incomes for people living in the United States better, about the same, or worse?
About United States (models are not told the place; the wording names it)
Wording “New York”
Compared with a month ago, are pay and incomes for people living in New York better, about the same, or worse?
About New York, USA (models are not told the place; the wording names it)
Wording “Pennsylvania”
Compared with a month ago, are pay and incomes for people living in Pennsylvania better, about the same, or worse?
About Pennsylvania, USA (models are not told the place; the wording names it)
Wording “Ohio”
Compared with a month ago, are pay and incomes for people living in Ohio better, about the same, or worse?
About Ohio, USA (models are not told the place; the wording names it)
Wording “Georgia”
Compared with a month ago, are pay and incomes for people living in Georgia better, about the same, or worse?
About Georgia, USA (models are not told the place; the wording names it)
Wording “Illinois”
Compared with a month ago, are pay and incomes for people living in Illinois better, about the same, or worse?
About Illinois, USA (models are not told the place; the wording names it)
Wording “Texas”
Compared with a month ago, are pay and incomes for people living in Texas better, about the same, or worse?
About Texas, USA (models are not told the place; the wording names it)
Wording “Colorado”
Compared with a month ago, are pay and incomes for people living in Colorado better, about the same, or worse?
About Colorado, USA (models are not told the place; the wording names it)
The numbers: latest run as a table
Model
Option
Share
95% range
Answers
Claude Haiku 5.5
Better
24%
13% to 39%
42
Claude Haiku 5.5
About the same
57%
42% to 71%
42
Claude Haiku 5.5
Worse
19%
10% to 33%
42
Claude Sonnet 5.5
Better
7%
2% to 19%
42
Claude Sonnet 5.5
About the same
86%
72% to 93%
42
Official data
Figures from their publishers, linked to this question, each as it had been published by the time shown, never as revised since. An official estimate is not a poll result, publishes no range, and the models were never shown it.
up $0.05, or 0.1% from August 2026: better (higher is better here)
Published by U.S. Bureau of Labor Statistics. An official estimate; no range published.
U.S. Bureau of Labor Statistics, Average Hourly Earnings of All Employees, Total Private [CES0500000003], retrieved from FRED, Federal Reserve Bank of St. Louis; https://fred.stlouisfed.org/series/CES0500000003.
Compared with a month ago, are economic conditions for people living in California better, about the same, or worse?
About California, USA (models are not told the place; the wording names it)
Wording “Alaska”
Compared with a month ago, are economic conditions for people living in Alaska better, about the same, or worse?
About Alaska, USA (models are not told the place; the wording names it)
Wording “Massachusetts”
Compared with a month ago, are economic conditions for people living in Massachusetts better, about the same, or worse?
About Massachusetts, USA (models are not told the place; the wording names it)
Wording “Virginia”
Compared with a month ago, are economic conditions for people living in Virginia better, about the same, or worse?
About Virginia, USA (models are not told the place; the wording names it)
Wording “Missouri”
Compared with a month ago, are economic conditions for people living in Missouri better, about the same, or worse?
About Missouri, USA (models are not told the place; the wording names it)
Wording “Minnesota”
Compared with a month ago, are economic conditions for people living in Minnesota better, about the same, or worse?
About Minnesota, USA (models are not told the place; the wording names it)
Answers offered, shuffled on every call
Better · About the same · Worse
Asked of
Claude Haiku 5.5, Claude Sonnet 5.5, GPT-5.6 Luna, Gemini 3.5 Flash-Lite: 3 times per wording each run, through each provider's API. Web search on for every model, up to 1 search per answer.
Claude Sonnet 5.5
Worse
12%
5% to 25%
42
GPT-5.6 Luna
Better
19%
10% to 33%
42
GPT-5.6 Luna
About the same
26%
15% to 41%
42
GPT-5.6 Luna
Worse
55%
40% to 69%
42
Gemini 3.5 Flash-Lite
Better
5%
1% to 16%
42
Gemini 3.5 Flash-Lite
About the same
74%
59% to 85%
42
Gemini 3.5 Flash-Lite
Worse
21%
12% to 36%
42
178.25, September 2026
up 0.17, or 0.1% from August 2026: better (higher is better here)
Compared with a month ago, are job prospects for people living in California better, about the same, or worse?
About California, USA (models are not told the place; the wording names it)
Wording “Alaska”
Compared with a month ago, are job prospects for people living in Alaska better, about the same, or worse?
About Alaska, USA (models are not told the place; the wording names it)
Wording “Massachusetts”
Compared with a month ago, are job prospects for people living in Massachusetts better, about the same, or worse?
About Massachusetts, USA (models are not told the place; the wording names it)
Wording “Virginia”
Compared with a month ago, are job prospects for people living in Virginia better, about the same, or worse?
About Virginia, USA (models are not told the place; the wording names it)
Wording “Missouri”
Compared with a month ago, are job prospects for people living in Missouri better, about the same, or worse?
About Missouri, USA (models are not told the place; the wording names it)
Wording “Minnesota”
Compared with a month ago, are job prospects for people living in Minnesota better, about the same, or worse?
About Minnesota, USA (models are not told the place; the wording names it)
Answers offered, shuffled on every call
Better · About the same · Worse
Asked of
Claude Haiku 5.5, Claude Sonnet 5.5, GPT-5.6 Luna, Gemini 3.5 Flash-Lite: 3 times per wording each run, through each provider's API. Web search on for every model, up to 1 search per answer.
Claude Sonnet 5.5
Worse
26%
15% to 41%
42
GPT-5.6 Luna
Better
50%
36% to 64%
42
GPT-5.6 Luna
About the same
14%
7% to 28%
42
GPT-5.6 Luna
Worse
36%
23% to 51%
42
Gemini 3.5 Flash-Lite
Better
10%
4% to 22%
42
Gemini 3.5 Flash-Lite
About the same
79%
64% to 88%
42
Gemini 3.5 Flash-Lite
Worse
12%
5% to 25%
42
up 0.1 points from August 2026: worse (lower is better here)
About Colorado, USA (models are not told the place; the wording names it)
Wording “California”
Compared with a month ago, are everyday prices for people living in California better, about the same, or worse for their household budgets?
About California, USA (models are not told the place; the wording names it)
Wording “Alaska”
Compared with a month ago, are everyday prices for people living in Alaska better, about the same, or worse for their household budgets?
About Alaska, USA (models are not told the place; the wording names it)
Wording “Massachusetts”
Compared with a month ago, are everyday prices for people living in Massachusetts better, about the same, or worse for their household budgets?
About Massachusetts, USA (models are not told the place; the wording names it)
Wording “Virginia”
Compared with a month ago, are everyday prices for people living in Virginia better, about the same, or worse for their household budgets?
About Virginia, USA (models are not told the place; the wording names it)
Wording “Missouri”
Compared with a month ago, are everyday prices for people living in Missouri better, about the same, or worse for their household budgets?
About Missouri, USA (models are not told the place; the wording names it)
Wording “Minnesota”
Compared with a month ago, are everyday prices for people living in Minnesota better, about the same, or worse for their household budgets?
About Minnesota, USA (models are not told the place; the wording names it)
Answers offered, shuffled on every call
Better · About the same · Worse
Asked of
Claude Haiku 5.5, Claude Sonnet 5.5, GPT-5.6 Luna, Gemini 3.5 Flash-Lite: 3 times per wording each run, through each provider's API. Web search on for every model, up to 1 search per answer.
Compared with a month ago, is housing affordability for people living in California better, about the same, or worse?
About California, USA (models are not told the place; the wording names it)
Wording “Alaska”
Compared with a month ago, is housing affordability for people living in Alaska better, about the same, or worse?
About Alaska, USA (models are not told the place; the wording names it)
Wording “Massachusetts”
Compared with a month ago, is housing affordability for people living in Massachusetts better, about the same, or worse?
About Massachusetts, USA (models are not told the place; the wording names it)
Wording “Virginia”
Compared with a month ago, is housing affordability for people living in Virginia better, about the same, or worse?
About Virginia, USA (models are not told the place; the wording names it)
Wording “Missouri”
Compared with a month ago, is housing affordability for people living in Missouri better, about the same, or worse?
About Missouri, USA (models are not told the place; the wording names it)
Wording “Minnesota”
Compared with a month ago, is housing affordability for people living in Minnesota better, about the same, or worse?
About Minnesota, USA (models are not told the place; the wording names it)
Answers offered, shuffled on every call
Better · About the same · Worse
Asked of
Claude Haiku 5.5, Claude Sonnet 5.5, GPT-5.6 Luna, Gemini 3.5 Flash-Lite: 3 times per wording each run, through each provider's API. Web search on for every model, up to 1 search per answer.
Claude Sonnet 5.5
Worse
36%
23% to 51%
42
GPT-5.6 Luna
Better
5%
1% to 16%
42
GPT-5.6 Luna
About the same
7%
2% to 19%
42
GPT-5.6 Luna
Worse
88%
75% to 95%
42
Gemini 3.5 Flash-Lite
Better
17%
8% to 31%
42
Gemini 3.5 Flash-Lite
About the same
57%
42% to 71%
42
Gemini 3.5 Flash-Lite
Worse
26%
15% to 41%
42
Wording “California”
Compared with a month ago, are pay and incomes for people living in California better, about the same, or worse?
About California, USA (models are not told the place; the wording names it)
Wording “Alaska”
Compared with a month ago, are pay and incomes for people living in Alaska better, about the same, or worse?
About Alaska, USA (models are not told the place; the wording names it)
Wording “Massachusetts”
Compared with a month ago, are pay and incomes for people living in Massachusetts better, about the same, or worse?
About Massachusetts, USA (models are not told the place; the wording names it)
Wording “Virginia”
Compared with a month ago, are pay and incomes for people living in Virginia better, about the same, or worse?
About Virginia, USA (models are not told the place; the wording names it)
Wording “Missouri”
Compared with a month ago, are pay and incomes for people living in Missouri better, about the same, or worse?
About Missouri, USA (models are not told the place; the wording names it)
Wording “Minnesota”
Compared with a month ago, are pay and incomes for people living in Minnesota better, about the same, or worse?
About Minnesota, USA (models are not told the place; the wording names it)
Answers offered, shuffled on every call
Better · About the same · Worse
Asked of
Claude Haiku 5.5, Claude Sonnet 5.5, GPT-5.6 Luna, Gemini 3.5 Flash-Lite: 3 times per wording each run, through each provider's API. Web search on for every model, up to 1 search per answer.