Hausman Test - Do I correct for heteroscedasticity and autocorrelation?

#1
Hi fella's!

I am currently working on a regression analysis of a demand curve and i'm trying to test endogeneity (because I expect the price to be related to the supply curve). I'm trying to test this via the Hausman test.

The problem is: I've found heterescedasticity in my model and until now I've fixed it using the White's heteroscedasticity-consistent errors and covariance.

I execute the hausman test, i.e. use OLS with the Price on all the exogenous variablles and then use OLS on my regular demand curve with an extra variable, the residuals of the previous OLS. I would then check the t-statistic to see if the coefficient is significantly different from 0.

Now my question: do I use the White's heteroscedasticity-consistent variance? And if yes, do I use it in the first or in the second OLS?

Thanks in advance!