Which is stupid. Everyone can be wrong because we haven't been wrong enough times to be right. How many people have to be wrong the same way before we benefit from the paths they re-tread?
Most research is funded by private corporations looking to make their money back then some. Two years of not being right is usually worse for having a sustainable career than just bsing that you're right. Case study: Alzheimer's research.
In my field they do publish results without success, but it must either be (a) something seminal in the field or (b) interesting in a notable way. General things aren't going to have the juice to get through the review process. One exception to this is the shotgun method. If you're testing a bunch of different things that get at the same question and they all miss, you might still get published, but that's because it's adjacent to (b).