But it is hard to remove even useless programs, as doing so launches a firestorm of protest from beneficiaires. As Wente says, programs are measured on volume of cash spent, not outcomes.
Now comes the discouraging part. The evidence to date – such as it is – suggests that many, perhaps most, social programs do not make a difference, except to the legions of administrators and social workers who are directly and indirectly employed in delivering them. This is not a conservative conclusion. It is the conclusion of independent groups such as the Brookings Institution (a non-partisan think tank) and the Coalition for Evidence-Based Policy, which are part of a growing movement to make social spending more accountable.
Here’s one example of how big the problem is, as laid out by the Brookings Institution. The U.S. federal government funds dozens of programs to help youth. These include a $1.2-billion after-school program for disadvantaged youth, a $1.5-billion Job Corps program for at-risk high-school students, and the legendary Head Start program, which spends $7-billion a year to help disadvantaged younger children. Ten of these programs, including Head Start, have been evaluated using the gold standard benchmark of random control groups. Nine of the evaluations found weak or no positive effects. A Brookings report says, “Only one program [Early Head Start, aimed at even younger children] was found to produce meaningful, though modest, positive effects.”
That does not mean all government spending is bad, as some Republicans think. But it does mean that if liberals persist in refusing to identify what does and does not work, or talking about "the vulnerable" in undifferentiated terms, they undermine the case for it. Good intentions are not enough: indeed, they can often be wholly counterproductive if there is no practical effort to deliver better outcomes.