Hi everyone, Is there a paper that talks about matching when the sample has more treatment observations than control observations? Is there an algorithm that works better for this case? Can somebody explain to me why optimal matching does not work at all in this case? Thanks, Ignacio