I am working on a project where the dependent variable and an independent variable are both occupation-level BLS estimates (employment and wages, respectively), and have associated percent relative standard errors (PRSEs): https://www.bls.gov/ecec/factsheets/...ard-errors.htm.
I'm vaguely familiar with generated regressor issues, though it doesn't seem possible to use any of the usually-recommended fixes (IV, bootstrap) with solely the estimated data and PRSEs. Are there ways to adjust for this in Stata?
I'm vaguely familiar with generated regressor issues, though it doesn't seem possible to use any of the usually-recommended fixes (IV, bootstrap) with solely the estimated data and PRSEs. Are there ways to adjust for this in Stata?
