survey of the effects of privatization

I found an interesting case study in a research method book by Bryman and Others (2011) were they were pointing out the effects of of privatization.  They describe the difficulties of obtaining access to some companies that were under the process of privatization due to sensitive information.  I find that if companies are able to filter out what information can be given, then the questions must be modified for it to be answered.  This was the case for the study because the researchers had to modify the questions which created non-sampling error.  In addition to that, the participants were chosen by the managements (human resource manager), so this reflected a bias of the company and created a source of sampling error.  I did not agree with the decision because the participants that were excluded were the manual workers (= low literacy).  Is this a socially acceptable barrier because it does bring up the issue of human rights and equality.  Overall, sampling errors and non-sampling errors sometimes do interrelate and it shows how non-response rates and sampling frames can be affected by the sensitivity of the issue.