This seems like a problem that appears everywhere in science, not just in the case of economics. There is a known phenomena of science results agreeing with the organizations that fund the science, and this seems to be an extension to that.
Is there any good way to solve it that won't result in less valid science work getting funded?
>Is there any good way to solve it that won't result in less valid science work getting funded?
Bona fide science has peer review and falsifiability. Also imperfect, but nowhere nearly as commonly manipulated as is economics for hire by business interests and their think tanks.
In fact, this seems like a good time to mention the only Senator I've ever seen in decades of observing politics who actively attacks this problem is Elizabeth Warren.
Funnel plots are used in meta-analysis to detect the self-censoring bias of publishing only results when an effect showed up. Maybe they can be used or adapted for this too?
Is there any good way to solve it that won't result in less valid science work getting funded?