> You've just eliminated all social sciences
Because they're indeed not sciences.
They call themselves that to attract funding.
Predictive power and reproducibility are defining characteristics of sciences.
Social "sciences" have neither and therefore aren't sciences.
reproducibility has improved over time as well - https://pubs.aeaweb.org/doi/pdfplus/10.1257/jel.20171350
edit: your take-away from this should be that you're mis-defining science.
edit2: you all forgot that medical and nutritional science has a reproducibility crisis, it's no less "science" for it. Add to the fact, the notion that nothing is reproducible in Economics is false. See for example: https://www.caltech.edu/about/news/famous-economics-experime... , https://www.science.org/doi/10.1126/science.aaf0918