everything peter hull: www.nber.org/papers/w27845

shift-share is the new black

without knowing any more info, for a statistical test, probably pearson or spearman (rank) correlation. but probably worth checking the scatterplot to see if there's interesting nonlinearity (usually more legible is binscatter, favorite package I use for this is seaborn.pydata.org/generated/se...)

If you want to support n+1 (thank you!), here are the links:

Donate: nplusonemag.com/donate
Subscribe: nplusonenag.com/subscribe
Gift sub: shop.nplusonemag.com/products/gif...
Buy merch or books: shop.nplusonemag.com