Hi:
By default, when you cross an analysis variable with a statistic, missing values are ignored in the calculation of statistics. I'm guessing that you have several observations with missing values for the "CreditHireTotal" variable.
If you run the code below, you will see that a copy of SASHELP.CLASS has been altered so that the value the SEX variable for 2 people has been changed and 2 heights have been turned to missing. The MISSING option reveals the 2 missing obs for the SEX variable, but has no effect on the count when it's based on (crossed with) height. The TABULATE program shows the difference between asking for the simple N and asking for the N crossed with the analysis variable.
cynthia
[pre]
data class;
set sashelp.class;
if name = 'Alfred' or name = 'Alice'
then sex = ' ';
else if _n_ gt 17 then height=.;
run;
proc print data=class;
title 'note missing values';
run;
proc freq data=class;
tables sex height / nocum nopercent;
run;
PROC TABULATE DATA=class missing;
CLASS sex;
VAR height;
TABLE sex='Gender' all,
height*(N='NbrOfStudents') / box='Ex 1';
table sex='Gender' all,
n='NbrOfStudents' / box='Ex 2';
table sex='Gender' all,
n='Total count' height*N='non missing heights'
/ box='Ex 3';
Title;
RUN;
[/pre]