Except in jurisdictions of mandatory voting, the vote is only ever by a portion of the electorate. Why then is there reporting on supposed approval after an election win? Gauging the mood of the broad population is not particularly important. Of course they try to predict likely voters and sample from that, but there always lingers some change of who actually votes in the next election being comparable to the last election. Trump's approval rating from 43% of Americans is possibly 60% of people who vote.
