No, what I meant is there is no need to use dynamic panel model. You still need to run Hausman test to decide which model (fixed vs. random) is more appropriate. In PLM package there are options for you to choose your models and tests (http://cran.r-project.org/web/packages/plm/vignettes/plm.pdf)