Accepted Paper
AI and Inequality paradox: when income concentration hinders AI scientific development in the North but drives it in the South
Susana Herrero
(Universitat de Barcelona)
Paper short abstract
The study of 86 countries (2000–2023) shows that in developed economies higher income inequality reduces AI scientific production, while in developing ones it increases it. Additionally, in developing countries, in the short term, higher scientific output tends to raise income inequality.
Paper long abstract
In this paper, we analyze the relationship between artificial intelligence (AI) scientific production and income inequality in a sample of eighty-six countries from 2000 to 2023. We estimate a PVAR model using the System-GMM approach and analyze the Impulse Response Functions (IRFs). Our results suggest that, in developed countries, a higher concentration of income is associated with a decline in AI scientific production. Meanwhile, in developing countries, the relationship is the opposite: greater income concentration is linked to higher AI scientific output. We also find that, in developing economies, in short-term, higher scientific production tends to be associated with higher income concentration.
The new South in global development