I agree with the former (GiveDirectly and other institutions like J-PAL have been huge in putting together randomized trials around aid effectiveness, [1] for a specific example). The latter I think is more up for debate. There are papers showing that aid is still effective in areas with high government corruption. Definitely a pretty debated area though in terms of effectiveness of foreign aid.

[1]. https://www.povertyactionlab.org/case-study/giving-directly-...