[gruez]

>This could, in turn, hinder the development of more effective policies and programs aimed at reducing poverty.

I don't buy this. If you want to figure out what the most effective policies/programs are, you'd need to run analysis on each program separately. If you did that, you would be able to separate out which parts had the greatest ROI. Dropping the most expensive component on the basis that it's ineffective is basically cherrypicking, especially when the claim that's being analyzed is whether government anti-poverty programs as a whole are effective or not.