I am using the clogit command in R, however I am not able to cluster the standard errors by districts using the method = "exact". It works with method = "error" but it gives me very different coefficients. Is there a way to cluster standard errors in the clogit command with the method = "exact"?

model1 <- clogit(mig ~ rainShock + factor(year) + strata(hhcode), data = data)
Error in clogit(mig ~ rainShock + factor(year) + strata(hhcode) + cluster(district),  : 
  robust variance plus the exact method is not supported