Nikoleta Anesti, Edward Hill, Andreas Joseph
arXiv 16 Dec 2025 · cs.CL
arXiv:2512.14306 · PDF · DOI · OpenAlex · Extracted main text
This paper investigates the ability of Large Language Models (LLMs), specifically GPT-3.5-turbo (GPT), to form inflation perceptions and expectations based on macroeconomic price signals. We compare the LLM's output to household survey data and official statistics, mimicking the information set and demographic characteristics of the Bank of England's Inflation Attitudes Survey (IAS). Our quasi-experimental design exploits the timing of GPT's training cut-off in September 2021 which means it has no knowledge of the subsequent UK inflation surge. We find that GPT tracks aggregate survey projections and official statistics at short horizons. At a disaggregated level, GPT replicates key empirical regularities of households' inflation perceptions, particularly for income, housing tenure, and social class. A novel Shapley value decomposition of LLM outputs suited for the synthetic survey setting provides well-defined insights into the drivers of model outputs linked to prompt content. We find that GPT demonstrates a heightened sensitivity to food inflation information similar to that of human respondents. However, we also find that it lacks a consistent model of consumer price inflation. More generally, our approach could be used to evaluate the behaviour of LLMs for use in the social sciences, to compare different models, or to assist in survey design.
appendix boundary found by appendix_titled_section at “Appendix” · 87% of the source is main text. Read the extracted text to check this.
The works this paper leans on most, across its whole bibliography — not restricted to papers in our corpus. Ranked by composite intensity, which combines how often a work is mentioned, how many sections mention it, and how much of that falls in the main text rather than the appendix.
| Reference | Intensity | Mentions | Sections | Main text | |
|---|---|---|---|---|---|
| 1 | D'Acunto, Francesco and Malmendier, Ulrike and Ospina, Juan and Webe… (2021) Exposure to grocery prices and inflation expectations | 0.737 | 3 | 2 | 100% |
| 2 | Argyle, Lisa P. and Busby, Ethan C. and Fulda, Nancy and Gubler, Jos… (2023) Out of One, Many: Using Language Models to Simulate Human Samples | 0.644 | 2 | 2 | 100% |
| 3 | Horton, John (2023) Large language models as simulated economic agents: What can we learn from homo silicus? | 0.644 | 2 | 2 | 100% |
| 4 | Strumbelj, Erik and Kononenko, Igor (2010) An efficient explanation of individual classifications using game theory | 0.644 | 2 | 2 | 100% |
| 5 | Arora, Arnav and Kaffee, Lucie-aimée and Augenstein, Isabelle (2023) Probing Pre-Trained Language Models for Cross-Cultural Differences in Values | 0.644 | 2 | 2 | 100% |
| 6 | Brown, Tom and Mann, Benjamin and Ryder, Nick and Subbiah, Melanie a… (2020) Language models are few-shot learners | 0.644 | 2 | 2 | 100% |
| 7 | Jens Ludwig and Sendhil Mullainathan and Ashesh Rambachan (2025) Large Language Models: An Applied Econometric Framework | 0.644 | 2 | 2 | 100% |
| 8 | Acemoglu, Daron and Autor, David and Hazell, Jonathon and Restrepo,… (2022) Artificial Intelligence and Jobs: Evidence from Online Vacancies | 0.405 | 1 | 1 | 100% |
| 9 | Aher, Gati and Arriaga, Rosa and Kalai, Adam Tauman (2022) Using large language models to simulate multiple humans | 0.405 | 1 | 1 | 100% |
| 10 | Bernanke, Ben (2007) Inflation expectations and inflation forecasting | 0.405 | 1 | 1 | 100% |
Showing the top 10 of 67 scored citations.