As a one-time PL/1 programmer (we got our record layouts from you COBOL guys), I'd eschew the record structure you want to use. ikegami gave you a great solution; I'd suggest tweaking

   my %Employee_Rec;
   @Employee_Rec{qw(
      Emp_No
      Emp_Lname
      Emp_Fname
      Emp_SSN
      Emp_DOB
      Emp_aka
   )} = $csv->fields();

   print "$Employee_Rec{Emp_No}     \n";
   print "$Employee_Rec{Emp_Lname}  \n";
   print "$Employee_Rec{Emp_Fname}  \n";
   print "$Employee_Rec{Emp_SSN}    \n";
   print "$Employee_Rec{Emp_DOB}    \n";
   print "$Employee_Rec{Emp_aka}    \n";
[download]

to

   my ( $Emp_No, $Emp_Lname, $Emp_Fname, $Emp_SSN, $Emp_DOB, $Emp_aka 
+) = $csv->fields();

   print "$Emp_No     \n";
   print "$Emp_Lname  \n";
   print "$Emp_Fname  \n";
   print "$Emp_SSN    \n";
   print "$Emp_DOB    \n";
   print "$Emp_aka    \n";
[download]

You gain nothing from storing the information in a hash (unless you were going to make a Hash of Hashes, keyed on the Employee Number). Prefacing the fields with Emp_ indicates the fields are related.