Potential disadvantages of vce(robust)

[Chris Martin]

I'm familiar with the purpose and reasoning behind vce(robust) but I would like to know if using vce(robust) with regress can ever cause any harm. In other words, are there ever disadvantages to using it?

[Rich Goldstein]

yes; e.g., if you are estimating what is the wrong model, then using robust SE's can hide a problem that you might otherwise notice; note that both David Freedman and Gary King have, separately, written papers about this; I am on vac and don't want to take the time to search but these should be findable

I found the Freedman paper anyway: Freedman, D (2006), " On The So-Called “Huber Sandwich Estimator” and “Robust Standard Errors” , The American Statistician, 60(4): 299-302

and the other paper: King, Gary, and Margaret Roberts. “How Robust Standard Errors Expose Methodological Problems They Do Not Fix, and What to Do About It,” Political Analysis, Vol. 23, No. 2 (2015): Pp. 159–179