Large Language Models (LLMs) as Agents for Augmented Democracy
This addresses the challenge of accurately capturing nuanced political preferences for researchers and policymakers, though it is incremental as it builds on existing LLM methods.
The study tackled the problem of predicting political preferences by using fine-tuned LLMs to augment citizen preference data from Brazil's 2022 election, finding that LLMs predicted individual choices more accurately than a baseline rule and improved aggregate preference estimates over non-augmented samples.
We explore an augmented democracy system built on off-the-shelf LLMs fine-tuned to augment data on citizen's preferences elicited over policies extracted from the government programs of the two main candidates of Brazil's 2022 presidential election. We use a train-test cross-validation setup to estimate the accuracy with which the LLMs predict both: a subject's individual political choices and the aggregate preferences of the full sample of participants. At the individual level, we find that LLMs predict out of sample preferences more accurately than a "bundle rule", which would assume that citizens always vote for the proposals of the candidate aligned with their self-reported political orientation. At the population level, we show that a probabilistic sample augmented by an LLM provides a more accurate estimate of the aggregate preferences of a population than the non-augmented probabilistic sample alone. Together, these results indicates that policy preference data augmented using LLMs can capture nuances that transcend party lines and represents a promising avenue of research for data augmentation.