It is possible to create
identical results with a basic DATA step match-merge and a PROC SQL
join. However, there are significant differences between these two
methods, as well as advantages and disadvantages to each. In some
cases, such as when there is a one-to-one or a one-to-many match on
values of the BY variables in the input data sets, these two methods
produce identical results. In other cases, such as when there is a
many-to-many match on values of the BY variables, or if there are
nonmatching values of the BY variables, these two methods produce
different results. These differences reflect the fact that the processing
is different for a DATA step match-merge and a PROC SQL join. Even
if you are working with many-to-many matches or nonmatching data,
it is possible to use other DATA step techniques such as multiple
SET statements to create results that are identical to the results
that a PROC SQL step creates.