Automated loan approvals erase gender bias in small business lending

A bank in Peru found that women were less likely to get loans and received smaller amounts when human loan officers made the decisions. When an automated system made the decisions, this bias disappeared. This means that using algorithms can make lending fairer for women entrepreneurs.

For years, people have worried that automated systems might just bake in existing human biases, making discrimination worse. This study shows the opposite: taking human discretion out of the loop can actually fix some of those biases. It suggests that if you want to make lending more equitable, you might need fewer people and more code.