What is the result of the DATA step when it specifies _NULL_ as the output dataset?

When the DATA step specifies NULL as the output dataset, the primary outcome is that no dataset is created. The value of NULL is used effectively to indicate that the programmer does not wish to create a new dataset but still wants to execute the DATA step code. This is often utilized when processing data is necessary, such as when performing calculations or generating reports, but storing the results in a permanent dataset is not desired.

In this context, the code within the DATA step may still run, and variables can be created and utilized, but they won't be saved into a new dataset. Instead, they may be used temporarily during the execution of the DATA step, and any manipulation will not generate any persistent output.

This approach is beneficial in scenarios where the intention is just to execute certain operations without the overhead of dataset creation. For instance, it can be used for logging messages, performing computations that are not required to be stored, or carrying out operations that have side effects without needing a dataset to hold their results.