12-29-2014 04:54 PM
I have unbalanced panel dataset N=2590, T=5 years (with gaps). I tried to estimate pooled OLS with HAC standard errors using proc panel. The results provide values for the parameter estimates but the standard errors are all zeros, t and P values are missing (.). The log does not provide any feedback. Any idea what is going wrong? Is it due to gaps in the data?