We can verify the number of values of salary_add by doing a little mathematics. As a Cartesian product takes place between all of the same index values, we can sum the square of their individual counts. Even missing values in the index produce Cartesian products with themselves:

>>> index_vc = salary1.index.value_counts(dropna=False)
>>> index_vc
Black or African American            700
White                                665
Hispanic/Latino                      480
Asian/Pacific Islander               107
NaN                                   35
American Indian or Alaskan Native     11
Others                                 2
Name: RACE, dtype: int64

>>> index_vc.pow(2).sum()
1175424