Richer Output for Richer Countries: Uncovering Geographical Disparities in Generated Stories and Travel Recommendations

TL;DR

This paper investigates how large language models encode geographical biases, revealing disparities in travel recommendations and story generation that favor wealthier countries and depict poorer regions negatively.

Contribution

It uncovers the impact of geographical biases in language models on real-world applications like travel and storytelling, highlighting disparities based on country wealth.

Findings

Travel recommendations for poorer countries are less diverse and contain fewer location references.

Stories from poorer regions more frequently express hardship and sadness.

Models encode geographical biases affecting output diversity and emotional tone.

Abstract

While a large body of work inspects language models for biases concerning gender, race, occupation and religion, biases of geographical nature are relatively less explored. Some recent studies benchmark the degree to which large language models encode geospatial knowledge. However, the impact of the encoded geographical knowledge (or lack thereof) on real-world applications has not been documented. In this work, we examine large language models for two common scenarios that require geographical knowledge: (a) travel recommendations and (b) geo-anchored story generation. Specifically, we study five popular language models, and across about $100$K travel requests, and $200$K story generations, we observe that travel recommendations corresponding to poorer countries are less unique with fewer location references, and stories from these regions more often convey emotions of hardship and sadness compared to those from wealthier nations.