Give groups larger grants for publishing negative results too? More money seems to be workable enough in the rest of the economy. Obviously it would be best if those at the top make the change and let it cascade down. But any private source of funds could also experiment with changing the grant model.