Learn English – “x and y are what drive them” vs “x and y are what driveS them”

The actual usage stats from the Corpus of Contemporary American English and the British National Corpus look as follows:

                  COCA    BNC

are what drive      10      0
are what drives      3      1
are what have       11      2
are what has         6      0
are what make       97      6
are what makes      25      3
are what tell        2      0
are what tells       0      0

So as you can see, American English rather clearly leans the plural way, while British English is either undecided, or the corpus size is simply too small to say anything definitely.

Personally, I do have to point out that the COCA results are actually at odds with the dialect I speak, where the variant with "what drive" strikes me as ungrammatical. (The antedecent of what might be plural, but the what itself is singular. All you need to do is to look at the what clause, an embedded question. And "*what drive them to do Z" is not a grammatical question in my dialect, or indeed for all I can tell in any variety of English. The question is, "what drives them to do it", and you are providing the answer to that. The "are" is a red herring, because it agrees with "X and Y" and is not part of the what clause at all.)

However, this kind of discrepancy between different dialects is nothing unusual, and there is even a name for this particular phenomenon: notional agreement. (For example, that's also the reason why we say "a lot of people are", even though "a lot" is clearly singular.)