A review of the history of efforts in the 1970s and 1980s to evaluate major social programmes in the USA sustains the proposition that over this period the American establishment of policy-makers, agency officials, professionals, and social scientists did not know how to design and implement social programmes that were minimally effective, let alone spectacularly so. Whilst local programmes, especially those run by locals, may succeed, two categories of federal programmes have failed conspicuously and consistently: large federal programmes intended to change behaviour (as opposed to providing commodities or cash), and any programme trying in particular to change the behaviour of a clientele that is not already socialized into norms of working-class or middle-class society. Despite the funds spent, it remains unclear how to ensure any significant behaviour change in significant proportions of the urban underclass.
It only rains when you have something planned.